Abstract: The disclosure provides a movement determination method, a movement determination device, and a computer-readable storage medium. The method includes the following. In response to determining that a user has presented an initial movement corresponding to a to-be-detected movement, an anchor joint point among multiple joint points of the user is decided based on the to-be-detected movement. A movable range of each of the anchor joint points is decided. In response to determining that each of the anchor joint points leaves the corresponding movable range, it is determined that the user has stopped performing the to-be-detected movement.

Abstract: A method and a device for predicting sports performance, and a computer readable storage medium are provided. The method includes: collecting motion data of a user in a process of the user performing a specific action through a wearable device; and in response to determining that the motion data conforms to action characteristics of the specific action, predicting at least one sports performance of the user performing at least one other action based on the motion data.

Abstract: The disclosure provides a movement determination method, a movement determination device, and a computer-readable storage medium. The method includes the following. In response to determining that a user has presented an initial movement corresponding to a to-be-detected movement, an anchor joint point among multiple joint points of the user is decided based on the to-be-detected movement. A movable range of each of the anchor joint points is decided. In response to determining that each of the anchor joint points leaves the corresponding movable range, it is determined that the user has stopped performing the to-be-detected movement.

Abstract: The invention provides a human body detection method, a human body detection device, and a computer readable storage medium. The method includes the following. A plurality of image frames related to a human body is obtained. A plurality of joint coordinates are detected in each image frame, and a plurality of specific image frames are accordingly found out. An image region height corresponding to the human body in each specific image frame is obtained. A first joint coordinate of a first joint in each specific image frame is obtained. A second joint coordinate of a second joint in each specific image frame is obtained. An actual length between the first joint and the second joint is estimated based on a height of the human body and the image region height, the first joint coordinate, and the second joint coordinate in each specific image frame.

Abstract: An automated inventory management method is provided. Historical sales states of multiple stores are received, and models of each store and each category of products are pre-trained by a pre-training module. States of each store and each category of products are obtained by a multi-store-multi-category training module, and horizontal and vertical relevance training based on the pre-trained models of each store and each category of products is conducted. Relevance between stores, horizontal relevance between categories of products, and vertical relevance between multiple stores and multiple categories of products are determined by a state analysis module, so that multiple stores and categories of products with high correlation are linked to modify expected sales of each store and each category of products. Orders for multiple categories of products in each store are placed and purchase volumes of multiple categories of products in each store are determined by an inventory decision module.

Abstract: An automated inventory management method is provided. A historical sale record is received, and a future sale of an item is predicted based on the historical sale record to obtain a simulation result of an expected sale state of the item in the next sale cycle. According to the historical sale record of full categories of items and the simulation result of the item in the next sale cycle, an initial weight of the pre-training model is trained and used as a weight of an inventory decision module, and a purchase order that meets the expected sale record of the item in the next sale cycle is automatically generated. A reward feedback is calculated according to a current sale record and an inventory volume of the item and a purchase order of the previous sale cycle and input into the inventory decision module to order the item.

Abstract: A project generating method comprises extracting a keyword from a plurality of text files with a specified category, determining whether the keyword is a theme, extracting a geographical name from the text files corresponding to the theme, determining whether to keep the theme according to an internet volume of the theme, filtering a plurality of review files from a review website according to the geographical name, calculating a first ratio therein to determining whether to keep the theme, and generating a project including the geographical name and the theme serving as a recommendation row.

Abstract: A reaction force detection system for treadmills and detection method thereof is disclosed, wherein the system detects the current data generated by a user running on a treadmill comprising a running belt, a motor and an electronic circuit device by means of a current sensor, and then determines a current peak based on the current data, such that a ground reaction force peak can be determined according to the acquired current peak. Therefore, the present invention allows to determine the ground reaction force of the human body during running through the changes of the motor current signals inside the treadmill, such that these parameters can be further presented, analyzed, and stored in order to provide more scientific feedbacks for running trainings.

Abstract: A fund tracking system, a fund tracking method and a graphic user interface are provided. The fund tracking method is used to track a target fund. The fund tracking method includes the following steps. Several ETF asset classes are obtained according to a fund benchmark index of the target fund. Several representative ETFs are obtained according to the ETF asset classes. A simulated investment portfolio is generated according to the representative ETFs. Whether the simulated investment portfolio meets a verification condition is verified. If the simulated investment portfolio meets the verification condition, the simulated investment portfolio is outputted.

Abstract: The present application discloses a trading decision generation system and method. The trading decision generation method includes: obtaining a market information; performing a market view generation module to generate a market view according to the market information; performing a state integration module to generate a state according to the market information, the market view and a trading information; performing a decision parameter generation module to generate a decision parameter according to the state; and performing a decision generation module to generate a decision according to the decision parameter.

Abstract: A project generating method comprises extracting a keyword from a plurality of text files with a specified category, determining whether the keyword is a theme, extracting a geographical name from the text files corresponding to the theme, determining whether to keep the theme according to an internet volume of the theme, filtering a plurality of review files from a review website according to the geographical name, calculating a first ratio therein to determining whether to keep the theme, and generating a project including the geographical name and the theme serving as a recommendation row.

Abstract: An investment strategy rule generation method including the following steps is provided. Firstly, an investment strategy rule generator generates an investment strategy model according to an investment history trend. Then, a total investment return of each of N candidate investment rules is obtained by the investment strategy rule calculator, wherein each of the N candidate investment rules includes a candidate market direction rule. The obtaining step includes: performing an investment simulation in each of multiple time points in a time window of a time interval. Then, the total investment return under the operation of the investment simulations over the time interval is calculated by the investment strategy rule calculator. Then, the candidate investment rule corresponding to the best of the total investment returns is used as an investment strategy rule of the investment strategy model by the investment strategy rule calculator.

Abstract: A reaction force detection system for treadmills and detection method thereof is disclosed, wherein the system detects the current data generated by a user running on a treadmill comprising a running belt, a motor and an electronic circuit device by means of a current sensor, and then determines a current peak based on the current data, such that a ground reaction force peak can be determined according to the acquired current peak. Therefore, the present invention allows to determine the ground reaction force of the human body during running through the changes of the motor current signals inside the treadmill, such that these parameters can be further presented, analyzed, and stored in order to provide more scientific feedbacks for running trainings.

Abstract: An investment strategy rule generation method including the following steps is provided. Firstly, an investment strategy rule generator generates an investment strategy model according to an investment history trend. Then, a total investment return of each of N candidate investment rules is obtained by the investment strategy rule calculator, wherein each of the N candidate investment rules includes a candidate market direction rule. The obtaining step includes: performing an investment simulation in each of multiple time points in a time window of a time interval. Then, the total investment return under the operation of the investment simulations over the time interval is calculated by the investment strategy rule calculator. Then, the candidate investment rule corresponding to the best of the total investment returns is used as an investment strategy rule of the investment strategy model by the investment strategy rule calculator.

Abstract: A decision factors analyzing device and a decision factors analyzing device for analyzing a plurality of decision factors which cause a product of a product type to be purchased are provided. The method includes identifying a plurality of product sequences corresponding to the product type from a plurality of browse history data and a plurality of purchase history data corresponding to a plurality of consumers of a consumer database, wherein each of the product sequences includes a unpurchased product and a purchased product; obtaining a feature sequence according to the produce sequences and a plurality of product information; training a regression model corresponding to the product type according to K decision factors of the feature sequence to obtain an optimized regression model, and obtaining K decision values respectively corresponding to the K decision factors according to the optimized regression model to generate a decision factor sequence corresponding to the product type.

Abstract: A method for finding crowd movements is provided. The method includes: collecting location data associated with multiple user devices, mining frequent patterns in the location data to generate multiple representative sequences, and grouping the representative sequences into clusters to find crowd movements according to sequence distances between the representative sequences. Each representative sequence includes at least one line segment between a starting location point and an end location point.

Abstract: A decision factors analyzing device and a decision factors analyzing device for analyzing a plurality of decision factors which cause a product of a product type to be purchased are provided. The method includes identifying a plurality of product sequences corresponding to the product type from a plurality of browse history data and a plurality of purchase history data corresponding to a plurality of consumers of a consumer database, wherein each of the product sequences includes a unpurchased product and a purchased product; obtaining a feature sequence according to the produce sequences and a plurality of product information; training a regression model corresponding to the product type according to K decision factors of the feature sequence to obtain an optimized regression model, and obtaining K decision values respectively corresponding to the K decision factors according to the optimized regression model to generate a decision factor sequence corresponding to the product type.

Abstract: A method for predicting user preference comprising: obtaining a users group with multiple users and a history shopping record associated with multiple goods; selecting a goods parameter, each goods having the goods parameter; selecting a plurality of first values of the goods parameter according to the history shopping record; determining a determining a representative users group with multiple representative users according to the history shopping record and the first values; calculating a correlation between a user and one of the representative users in the representative users group; and predicting the preference of the users according to the correlation and representative users' preference.

Abstract: A method for finding crowd movements is provided. The method includes: collecting location data associated with multiple user devices, mining frequent patterns in the location data to generate multiple representative sequences, and grouping the representative sequences into clusters to find crowd movements according to sequence distances between the representative sequences. Each representative sequence includes at least one line segment between a starting location point and an end location point.

Abstract: According to one exemplary embodiment, a method for analyzing root causes applies an application-level dependency discovery and anomaly detection to find application-level dependencies in one or more virtual machines (VMs), and generate an application-level topology with anomaly, and then transfers the application-level topology with anomaly to a VM-level dependency, and transfers the VM-level dependency to a physical machine level (PM-level) dependency via a physical and virtual resource mapping, and eventually generates a group of event sets. A prioritized event list is generated by prioritizing the group of event sets.