Abstract: A plurality of potential fault scenarios are accessed, wherein a given potential fault scenario of the plurality of potential fault scenarios has at least one corresponding root cause, and a representation of the given potential fault scenario comprises a don't care value. An actual fault scenario from telemetry received from a monitored system is generated. The actual fault scenario is matched against the plurality of potential fault scenarios. One or more matched causes are output as one or more probable root cause failures of the monitored system.

Abstract: A method for checking policy compliance of events of an event stream includes receiving the events; grouping a plurality of the received events into a plurality of slices based upon a policy specification and an event classification; determining whether a policy violation has occurred by concurrently evaluating at least two of the slices according to the policy specification; and in a case in which the policy violation is determined, reporting the policy violation.

Abstract: The route prediction system according to the invention includes a measurement unit to measure an area including a host vehicle and other moving vehicles, a vehicle detection unit to detect the host vehicle and at least two of the surrounding vehicles having collision possibilities on the basis of observation results observed by the observation unit, a hypothesis generation unit to generate plural hypotheses for the at least two of the surrounding vehicles detected by the vehicle detection unit to avoid collision, a likelihood calculation unit to calculate a likelihood indicating probability of occurrence of each of the plural hypotheses generated by the hypothesis generation unit, and a predicted route analysis unit to analyze, on the basis of the likelihood calculated by the likelihood calculation unit, predicted routes of the at least two of the surrounding vehicles, and output the analysis result.

Abstract: A guard dog station for preventing entrance to a venue of individuals identified as not eligible to enter. The guard dog station has at least one camera extending to a height of the face of an average person. A processor within the guard dog station is connected to the at least one camera. The processor being connected to a database that contains images and names of individuals who are authorized or not authorized to enter the venue. A speaker mounted on the guard dog station produces the sound of a barking dog in response to identification of an unauthorized individual. Additionally provided is a laptop for the guard to view images of individuals approaching the guard dog station who are or are not authorized to enter the venue.

Abstract: A relationship relating a load of exercise and a user's aerobic capacity may be determined as follows. A processor circuit of a device may retrieve, from a memory, a prior probability distribution of the load of exercise and a prior probability distribution of the user's aerobic capacity. The processor circuit may compute a joint prior probability of the load of exercise and the user's aerobic capacity. The processor circuit may compute a joint likelihood of the load of exercise and the user's aerobic capacity based on data indicative of a measured time-stamped work rate and a measured time-stamped heart rate. The processor circuit may combine the joint prior probability and the joint likelihood to produce a joint posterior probability. The processor circuit may use the joint posterior probability to determine a relationship relating the load of exercise and the user's aerobic capacity and output a calorie calculation.

Abstract: Embodiments for facility management by a processor. Operations are performed to collect and track environmentally-related data of the facility over time. Predictive analytics are applied to the environmentally-related data to generate decisions that are dynamically applied to environmental settings of the facility.

Abstract: Some embodiments provide methods, systems and computer-readable media that employ adaptive binning and dissimilarity scores based on a quadratic form distance for multidimensional data for matching clusters in data corresponding to different sample. Some embodiments provide methods, systems and computer-readable media for rendering a first interactive display including a two-dimensional plot of at least a portion of a multidimensional data set and a corresponding second interactive display including a plurality of single parameter charts or histograms, each displaying information corresponding to one-dimensional measurements of a different parameter in the multidimensional data set.

Abstract: An event analysis system receives events in a time-series from a set of monitored systems and identifies a set of alert threshold values for each of the types of events to identify outliers in the time-series at an evaluated time. Portions of historic event data is selected to identify windows of event data near the evaluated time at a set of seasonally-adjusted times to predict the value of the event type. The alert threshold value may also account for a prediction based on recent, higher-frequency events. Using the alert threshold values for a plurality of event types, the event data is compared with the alert threshold values to determine an alert level for the data. The event data types are also clustered and displayed with the alert levels to provide a visualization of the event data and identify outliers when the new event data is received.

Abstract: An array of resistive processing units (RPUs) comprises a plurality of rows of RPUs and a plurality of columns of RPUs wherein each RPU comprises an AND gate configured to perform an AND operation of a first stochastic bit stream received from a first stochastic translator translating a number encoded from a neuron in a row and a second stochastic bit stream received from a second stochastic translator translating a number encoded from a neuron in a column. A first storage is configured to store a weight value of the RPU, and a second storage is configured to store an amount of change to the weight value of the RPU. When the first stochastic bit stream and the second stochastic bit stream coincide, the amount of change to the weight value of the RPU is added to the weight value of the RPU.

Abstract: A remote pump manager is provided. The remote pump manager includes pressure sensors or a depth sensor operating to determine the change in pressure of a pump, a flow sensor operating to determine flow rate of fluid exiting the pump, a power meter operating to determine power data related to operation of the pump, and a management device having a control device and a display. The pressure sensors, the flow sensor and the power meter are in communication with the management device. The management device operates to determine pump efficiencies, wherein the control device automatically determines pump efficiency data in response to receiving real time data from the pressure sensors, the flow sensor and the power meter and automatically delivers the pump efficiency data to the display for displaying the determined pump efficiency data.

Abstract: Provided is a plant state monitoring method which monitors an operation state of a plant by using the Mahalanobis distance based on the plant state amount. The method creates a first unit space as a set of data used to be a reference when judging whether the plant operation state during a start operation period is normal according to the state amount in the plant start operation period. The method also creates a second unit space as a set of data used to be a reference when judging whether the plant operation state during a load operation period is normal according to the state amount in the plant load operation period.

Abstract: An automated system and method provides for planning and response to disparate threats with combined kinetic (e.g., missile interceptor) and non-kinetic (e.g., cyber and electronic warfare) effects. The threat and effects inputs and the results of analysis tools that individually apply a single effect type to negate a threat vulnerability are normalized to a form and format ingestible by a Stochastic Math Model (SMM). The normalized inputs reflect the success/failure of the effect versus a threat vulnerability as a score, not a probability. The normalized results decompose a probability of defeat (Pdefeat) for an individual effect as a set of hierarchical parameters represented as probabilities.

Abstract: This disclosure relates to processing a spectral dataset, such as a hyperspectral image or a large collection of individual spectra taken with the same spectrometer, to increase the signal-to-noise ratio. The methods can also be used to process a stack of images that differ by acquisition time rather than wavelength. The methods remove most of the sensor background noise with minimal corruption of image texture, anomalous or rare spectra or waveforms, and spectral or time resolution.

Abstract: A battery's state of charge estimation apparatus will be provided. The battery's state of charge estimation apparatus includes: a charge and discharge current detection unit; a terminal voltage detection unit; an open circuit voltage method state of charge estimation unit for estimating an open circuit voltage of the battery and an open circuit voltage method state of charge; a current integration method state of charge estimation unit for obtaining a current integration method state of charge as a state variable; and an error correction value calculation unit for calculating an error correction value for correcting the current integration method state of charge. The current integration method state of charge estimation unit corrects the current integration method state of charge by using the error correction value.

Abstract: Systems and methods for modeling and forecasting cyclical demand systems in the presence of dynamic control or dynamic incentives. A method for modeling a cyclical demand system comprises obtaining historical data on one or more demand measurements over a plurality of demand cycles, obtaining historical data on incentive signals over the plurality of demand cycles, constructing a model using the obtained historical data on the one or more demand measurements and the incentive signals, wherein constructing the model comprises specifying a state-space model, specifying variance parameters in the model, and estimating unknown variance parameters.

Abstract: A system and method are disclosed for the collection and aggregation of data from contributing members of a community, such as health-related, personal, genomic, medical, and other data of interest for individuals and populations. Contributors become members of a community upon creation of an account and providing of data or files. The data is received and processed, such as to analyze, structure, perform quality control, and curate the data. Value or shares in one or more community databases are computed and attributed to each contributing member. The data is controlled to avoid identification or personalization. Steps are taken to determine incompleteness and incorrectness of the data, and the data may be improved or completed automatically, based upon interaction with members, additional contributions of data, and so forth.

Abstract: A method for defect pre-warning of a power device may comprise processing measurement data regarding components of the power device based on predetermined severity criteria, to obtain severities of monitoring indexes related to defects of the components. Weights of the monitoring indexes with respect to the defects are determined based on effectiveness of the monitoring indexes for the defects. A defect probability that at least one of the components has each of at least one of the defects is estimated based on the severities and the weights. Some embodiments may achieve more reliable defect probability estimation and in some embodiments defective component probability and/or device healthy may be estimated. Based on the reliable information, it may facilitate to arrange the maintenance before defects involve into a fault and assess a risk of power device operation and control, thereby improving reliability of the power device.

Abstract: A structural health monitoring system comprising intelligent sensors and serial communications devices interconnecting the sensors, data acquisition units and a programmable logic controller (PLC). The intelligent sensors are individually programmable and can be tuned to listen to specific frequencies based on a scaling factor based on the sensor's location in the system being monitored, a set of pre-known frequencies based on the sensor's location in the system being monitored, and the output of a central frequency-based system sensor. An aircraft flight control application of a method embodying the present invention includes anomaly detection based on comparing one or more preprogrammed flight plan regimen variables with actual aircraft performance parameters, and taking appropriate corrective action.

Abstract: A computer implemented method and system for determining reliability of a machine includes receiving one of a machine data from one or more locations through an internet of things (IOT) based machine wearable sensor network. The method further includes storing the data in a distributed computer database communicatively coupled to an enterprise resource planning (ERP) system and extracting, through a computer server, one or more entity information from the data to compare against a pre-defined baseline. Further, mapping, though a big data machine learning engine, the extracted one or more entity information into a multi-classification model. The method includes indicating, through a machine learning engine coupled to a predictive analytics engine, on a user interface a set of analytical predictions for machine maintenance, repair and operation.

Abstract: A system and method of obtaining an abstracted issue from a problem associated with a system include receiving a problem report. The problem report includes one or more symptoms of the problem. The method also includes deleting at least an identifier of the problem in the problem description to generate the abstracted issue, and storing the abstracted issue in a searchable database.

Abstract: Example implementations described herein are directed to systems and methods for defect rate analytics to reduce defectiveness in manufacturing. In an example implementation, a method include determining, from data associated with each feature for a manufacturing process, the data feature indicative of process defects detected based on the feature, an estimated condition for the feature that reduces a defect rate of the process defects, the estimated condition indicating the data into a first group and second group; calculating the rate reduction of the defect rate based on a difference in defects between the first group and the second group; for the rate reduction meeting a target confidence level for a target defect rate, applying the estimated condition to the manufacturing process associated with each of the features. In example implementations, the defect rate analytics reduce defectiveness in manufacturing with independent processes and/or dependent processes.

Abstract: Techniques are described for automatic interval metadata determination for intermittent time series data. In one example, a method for determining intermittent time series interval metadata includes detecting one or more time variables in a time series data set. The method further includes determining whether the one or more time variables are intermittently regular. The method further includes determining one or more respective time intervals for the one or more time variables. The method further includes determining the parameters of intermittency for the one or more time variables. The method further includes generating an output comprising information about the one or more time variables based on the one or more respective time intervals and the parameters of intermittency for the time variable.

Abstract: Techniques are described for automatic interval metadata determination for intermittent time series data. In one example, a method for determining intermittent time series interval metadata includes detecting one or more time variables in a time series data set. The method further includes determining whether the one or more time variables are intermittently regular. The method further includes determining one or more respective time intervals for the one or more time variables. The method further includes determining the parameters of intermittency for the one or more time variables. The method further includes generating an output comprising information about the one or more time variables based on the one or more respective time intervals and the parameters of intermittency for the time variable.

Abstract: A mechanism is provided for computing a solution to a plan recognition problem. The plan recognition problem includes the model and a partially ordered sequence of observations or traces. The plan recognition is transformed into an AI planning problem such that a planner can be used to compute a solution to it. The approach is general. It addresses unreliable observations: missing observations, noisy observations (or observations that need to be discarded), and ambiguous observations). The approach does not require plan libraries or a possible set of goals. A planner can find either one solution to the resulting planning problem or multiple ranked solutions, which maps to the most plausible solution to the original problem.

Abstract: A system includes a processor configured to wirelessly receive first usage data in conjunction with a wear-state report, indicating a sensed level of system wear, from a plurality of vehicles. The processor is also configured to aggregate and analyze the first usage data to determine common parameters and corresponding values indicative of the sensed level of system wear. The processor is further configured to wirelessly receive second usage data from a vehicle lacking a sensor capable of sensing the sensed level of system wear. Also, the processor is configured to compare values of the common parameters in the second usage data to the determined values indicative of the sensed level of system wear and report a likely wear-state to the vehicle lacking the sensor, responsive to the comparison indicating a level of system wear similar to the sensed level of system wear.

Abstract: A method for selecting advertisements to embed in an online application, including detecting an access to the online application by a customer, determining, based on an impression tracker associated with the customer, and using an ad type selection model, an advertisement to embed in the online application. The advertisement is displayed to the user and the impression tracker associated with the customer is incremented. The method further includes receiving, from the user, a selection of the advertisement, storing a type of the advertisement and the value of the impression tracker as ad selection data, determining that the ad type selection model needs to be updated, and updating the ad type selection model based on the ad selection data to obtain an updated ad type selection model.

Abstract: The disclosure relates to a method for determining the accuracy of a satellite-based navigation system for a vehicle during operation on a test route. The method includes providing a light beam propagating essentially perpendicularly to the direction of travel along the test route at each of two selected measurement points on the test route. Upon detection of the light beams impacting the vehicle perpendicularly to the direction of travel, a detection of the position and the direction of travel measured by a receiver of the satellite-based navigation system is triggered by means of an optical triggering device fastened on the vehicle. Finally, an accuracy of the satellite-based navigation system is determined on the basis of the two detected positions and directions of travel and position coordinates of the two selected measurement points. The disclosure also relates to a corresponding system for determining the accuracy of a satellite-based navigation system.

Abstract: Mapping anonymous Internet entities to known accounts. In an embodiment, events, representing online activity and comprising IP addresses, are received from a plurality of sources. Subsets of the events are aggregated into mappings that associate the IP address, shared by the subset, with an account. Each mapping is associated with statistics regarding the events. A confidence value is calculated for each mapping based on the statistics, and a final subset of the mappings is selected based on the confidence values. Subsequently, when a request with an IP address is received, the final subset of mappings is searched for the requested IP address, and an indication of the account associated with the requested IP address is returned in response to the request.

Abstract: Described is a system for identifying communication behavior patterns in communication activity time series. For each pair of variables in the communication activity time series, the system determines a transfer entropy measure, an effective transfer entropy measure from a randomly reordered version of the communication activity time series, and a partial effective transfer entropy measure. A dependency matrix is generated using pair-wised effective transfer entropy measures and partial effective transfer entropy measures, where each element in the matrix represents a total influence of a communication activity time series on another communication activity time series in the future. The dependency matrix is compared with dependency matrices generated from a predefined set of communication patterns to identify the communication behavior pattern.

Abstract: A diagnostics and prediction system including a cloud system that continuously collects operating parameters from each of a number of environmental sensors and provides access to this data by a plurality of processing applications including (1) a predictive modeling system including (a) a health prediction system, (b) a sensor false alarm prediction system, (c) a zone false alarm prediction system and (d) a reporting system, (2) a system that diagnoses and predicts environmental hazardous areas and clusters areas based upon concentrations of CO in the site or building; and (3) a battery prediction system that predicts a battery life for the sensor.

Abstract: A method for analyzing financial data generated in an exchange in communication with a client/server network. The method includes the steps of receiving a first set of data related to a financial instrument on the exchange, the first set of data being within a time interval and including: a high value for the financial instrument during the time interval; a low value for the financial instrument during the time interval; an opening value for the financial instrument during the time interval; and a closing value for the financial instrument during the time interval. The method also detects a spike in the time interval using a confidence value based upon the first set of data and displays a graph of a set of transactions of the financial instrument, the graph being scaled to prevent the spike from substantially affecting visual information relating to the set of transactions.

Abstract: Described herein is a system that transmits and combines local models, that individually comprise a set of local parameters computed via stochastic gradient descent (SGD), into a global model that comprises a set of global model parameters. The local models are computed in parallel at different geographic locations along with symbolic representations. Network transmission of the local models and the symbolic representations, rather than transmission of the large training data subsets processed to compute the local models and symbolic representations, conserves resources and decreases latency. The global model can then be used as a model to determine a likelihood of a monitored resource or a user of the monitored resource experiencing a problem with respect to performance or completion of one or more operations. The system can also implement an action to assist in resolving or avoiding the problem.

Abstract: Instead of accepting competing ads and using an arbitration function (e.g., an auction) to choose winning ads to be served with a document, sets of ads (perhaps having different characteristics) can be generated, and an arbitration function can be used to select the winning set of ads. Such arbitrations on sets of ads can consider how ads, search results, colors, positions, fonts, etc., all interact with each other and affect the usefulness of the sets of ads to advertisers, end users, document publishers, and/or an ad serving entity.

Abstract: A system may process and display communications data to a user. The system may receive data related to a time series of communication moments. The communication moments may include a property of a communication event that was derived based on an analysis of the communication event. The system may process the data to generate aggregate type information corresponding to one or more types of the communication moments. The system may display the communication moments and the aggregate type information to a user.

Abstract: Generating a contributor-based power abstract for a device, including: identifying a clock power component for each of a plurality of clock gating domains, identifying a switching characteristic for each of the clock gating domains, combining the switching characteristics for all of the clock gating domains into a domain combination list, performing a per-case simulation based at least on the domain combination list, calculating an effective capacitance for each of the clock gating domains based at least on the per-case simulation, and generating a power abstract for each of the clock gating domains based at least on the effective capacitance.

Abstract: Predictor variables that affect a building energy load are sampled at a plurality of times within a time period and aggregated to generate an aggregated value for each predictor variable over the time period. A model is generated which estimates the building energy load in terms of the predictor variables. A regression process is performed to generate values for a plurality of regression coefficients in the model based on a cumulative building energy load for the time period and the aggregated values. The sampled values of the predictor variables are then applied as inputs to the model to estimate building energy loads at each of the plurality of times. The estimated building energy loads may be used as inputs to a controller that optimizes operation of a central plant.

Abstract: An approach for the analysis of the impact of over the top (OTT) applications is provided, which operate “over the top” of the network applications, with respect to network performance and power consumption on mobile devices. A wireless network for an application-carrying device separate from wireless communications network test equipment is simulated. A wireless communications interface between the simulated network and the application-carrying device is monitored. The monitoring comprises measuring signaling data on a control plane, and measuring application data on a user plane related to at least one application implemented on the application-carrying device. Cross-layer measurement data is generated based on the measured signaling data and application data. The signaling data and the application data are measured on the basis of a common time reference of the test equipment.

Abstract: Embodiments of the present disclosure relate to a method and apparatus for managing a failure of a device. The method comprises detecting whether a failure occurs in a device, and generating a failure report for the failure in response to the failure occurring in the device. The method further comprises querying a device object repository with the failure report, and the object device repository stores historical failure information associated with the device and a fix solution corresponding to the historical failure information. The method further comprises obtaining the fix solution from the device object repository based on a comparison between the failure report and the historical failure information. Embodiments of the present disclosure can manage the failure of the device more effectively.

Abstract: Systems and methods for interactive emergency response systems are provided. Some methods can include retrieving a building information model of a monitored region stored in a database device, receiving real time data related to an occupant in the monitored region, receiving real time ambient condition signals from a plurality of devices in the monitored region that identify current conditions in the monitored region, dynamically generating and updating an evacuation plan for the occupant from the monitored region based on the building information model, the real time data, and the real time ambient condition signals, and dynamically transmitting a notification of the evacuation plan as updated to a user device carried the occupant for use in exiting the monitored region and moving to a safe assembly area.

Abstract: A computer system is monitored for conditions bearing on the effective age of various hardware components of the system. Upon detecting an age adjustment condition, the affected hardware components are identified. An age adjustment is determined for each of the affected hardware components. The adjusted age is stored for reference when taking actions based on the adjusted age, such a component replacement and/or workload placement.

Abstract: A method and system for dynamic locating of a fault observed in a defective component related to a forging operation, in which: a processor models a shaping operation of a component by forging, as per a set of successive models of the component; a processor adds a fault plotter to a first model of the set of models, in a zone corresponding to a region of the fault in the defective component to obtain a first plotted model; and a processor tracks the plotter in time from the first plotted model during the modelling to locate an origin of the fault.

Abstract: A method for controlling an electronic device includes determining a software application that is running in the electronic device. Pressing information in relation to a predetermined physical button of the electronic device is detected and a probability of use of each of buttons corresponding to the software application is obtained by searching a preset list according to the software application. An operation is performed according to the pressing information and the probability of use of each of buttons corresponding to the software application.

Abstract: A method of health management of a monitored system includes collecting component condition indicator data used to calculate a plurality of component health indicators. Component fault severity and potential failure modes are determined utilizing the component condition indicator data. The potential failure modes are ranked in order of likelihood to isolate the failure mode. A system of health management for monitored apparatus includes a fault severity module to derive a plurality of component health indicators from collected component condition indicator data, the plurality of component health indicators indicative of fault severity of a plurality of components. A fault isolation module separately derives a ranked listing of potential fault/failure modes utilizing the component condition indicator data.

Abstract: The present invention builds Home Data Index (HDI) models. One driving force behind this HDI initiative is that no one model or measure can truly capture the widely dynamic movement of home prices. Even within a small geographical area, such as a ZIP code, there is significant variation in property sale types, sale frequencies and sale values. To better describe these variations, the present invention presents a suite of paired sales and price per square foot index models built around an array of property transaction characteristics. These HDI models expand on the usage and understanding of traditional home price indices (HPIs) by implementing a multidimensional index comprised of four main dimensions: geography; time frames; value range; and property sales type. Several layers exist within each dimension, allowing for more than 300 different index model perspectives for a given property address.

Abstract: A mechanism is provided in a computing device configured with instructions executing on a processor of the computing device to implement a timeline generation system, for automated timeline completion. The timeline generation system executing on the processor of the computing device identifies a plurality of events in documents in a corpus of information. The timeline generation system places the plurality of events in a partial timeline data structure. The timeline generation system selects an event progression from an event progression knowledge base. The timeline generation system aligns the selected event progression to the partial timeline data structure. The timeline generation system identifies a set of events missing from the partial timeline data structure. The timeline generation system maps the set of events missing from the partial timeline data structure to the partial timeline based on the selected event progression to form a completed timeline data structure.

Abstract: Embodiments include methods, systems, and apparatuses for a first wireless transmit/receive unit (WTRU) to perform device-to-device (D2D) communication directly with a second WTRU is disclosed. The embodiments may include determining a base pattern length to use for the D2D communication based on at least a transmission configuration. A base pattern may be selected from one or more base patterns having the determined base pattern length based on an indication from a network. The selected base pattern may have one or more transmission opportunities. One or more data packets may be transmitted to the second WTRU in at least one of the one or more transmission opportunities of the selected base pattern.

Abstract: Systems and methods are provided for generating and implementing data auditing functionality supporting multiple database platforms. A database management computer system can engage a shadow audit module to automate and optimize the implementation of data auditing functionality for multiple database platforms that may be utilized by an enterprise application to record data interactions (e.g., insertions, updates, deletions, etc). The shadow audit module may run generated audit SQL code to create an audit trigger in the selected main data tables. Each audit trigger may be a procedure and/or a flag that is stored in the main data table to run or transmit a signal indicating a change may have occurred on the record indicating the required replication of the modification in the shadow audit data table.

Abstract: Methods for integrated circuit design are provided. In one embodiment, a method for determining a physical layout pattern includes accessing a layout pattern configuration graph. The graph includes layout pattern configurations meeting a circuit requirements. At least two of the layout pattern configurations are annotated with characteristics by analyzing sample layout patterns. An integrated circuit electrical design is partitioned into circuit design configurations. One of the circuit design configurations meets one of the circuit requirements. One of the layout pattern configurations is selected from the layout pattern configuration graph to meet the selected circuit requirements. In another embodiment, a method for determining a netlist for an integrated circuit electrical design is provided. In a further embodiment, a method for determining a tool configuration for a manufacturing process is provided.

Abstract: Presented are a method, apparatus, and computer-readable medium for aiding emergency response. The method includes receiving, by a processor, a plurality of data corresponding to a geographic location from a plurality of data sources. The method further includes aggregating, by the processor, the plurality of data. The method still further includes determining, by the processor, a first profile and a second profile based on the aggregated plurality of data. The method can conclude with transmitting, by the processor, the first profile to a plurality of user equipments and the second profile to at least one user equipment, wherein the at least one user equipment is different from the plurality of user equipments.

Abstract: A social networking system dynamically adjusts a number of advertisements presented to a user along with organic content items by modifying a minimum bid amount associated with advertisements eligible for presentation to the user. Increasing the minimum bid amount decreases the number of advertisements presented to the user while decreasing the minimum bid amount increases the number of advertisements presented to the user. An engagement score measuring the user's estimated interaction with a content feed including organic content items without advertisements and an engagement score measuring the user's estimated interaction with a content feed including organic content items and advertisements are determined. A target score is determined based on the engagement scores, and a difference between the target score and a threshold value is used to modify a minimum price of advertisements eligible for presentation to the user.