Abstract: A computer-implemented method comprising: obtaining an output from each of a plurality of kernels in an extraction layer of a first trained convolutional neural network, wherein the first convolutional neural network is configured to identify one or more features in an image; aggregating the outputs corresponding to at least some input samples of a first domain to generate an aggregate map corresponding to that kernel; resizing the aggregate maps to a lower resolution to generate a plurality of region maps corresponding to the aggregate maps, respectively; clustering the region maps to generate clusters of region maps, each cluster comprising region maps having similar regions; and training, using input samples of a second domain, a second convolutional neural network with a kernel weight of at least one of the kernels which corresponds to at least one of the image regions of at least one of the clusters.

Abstract: A system may include a rental property listing portal and a rental price prediction unit. The rental price prediction unit may additional include an amenity price prediction unit. The rental price prediction unit and the amenity price prediction unit may use one or more mathematical models, such as constrained optimization model and/or a machine learning approach such as a random forest model, to generate a suggested rental price at a given level of portfolio occupancy. The system may also provide risk thresholds to simulate leasing strategies based on market situations, predicting revenue potential and possible vacancy for each strategy. The system may also generate a delist time (time to rent) for a property at a given price. The system and method may be implemented in a multi-family apartment rental system, a storage space rental system, a student accommodation, and single family apartment rental systems.

Abstract: A system may include a rental property listing portal and a rental price prediction unit. The rental price prediction unit may additional include an amenity price prediction unit. The rental price prediction unit and the amenity price prediction unit may use one or more mathematical models, such as constrained optimization model and/or a machine learning approach such as a random forest model, to generate a suggested rental price at a given level of portfolio occupancy. The system may also provide risk thresholds to simulate leasing strategies based on market situations, predicting revenue potential and possible vacancy for each strategy. The system may also generate a delist time (time to rent) for a property at a given price. The system and method may be implemented in a multi-family apartment rental system, a storage space rental system, a student accommodation, and single family apartment rental systems.

Abstract: A system and method for provisioning wireless devices such that they are only capable of accessing network services when in a specified location. A device surveys the current location of the device for existing base stations and stores a corresponding network environment profile. After the profile is stored, network services will only be provided to the device if the current network profile matches the stored profile.

Abstract: A system and method for provisioning wireless devices such that they are only capable of accessing network services when in a specified location. A device surveys the current location of the device for existing base stations and stores a corresponding network environment profile. After the profile is stored, network services will only be provided to the device if the current network profile matches the stored profile.

Abstract: A system and method for provisioning wireless devices such that they are only capable of accessing network services when in a specified location. A device surveys the current location of the device for existing base stations and stores a corresponding network environment profile. After the profile is stored, network services will only be provided to the device if the current network profile matches the stored profile.

Abstract: A mobile hotspot device may be used to interface one or more host computing devices with a wireless wide area network. The mobile hotspot device is powered by a high-capacity lithium ion (Li-Ion) cylindrical battery pack(s) and has a small, pen-like form factor. Use of the Li-Ion cylindrical battery pack(s) improves mobile hotspot device talk time over prismatic battery cell technology, as well as avoids premature device/modem shutdown experienced when a prismatic battery cell is proximate to a printed circuit assembly of the mobile hotspot device. Additionally still, Li-Ion cylindrical battery pack(s) avoid negatively impacting volume capacity of the mobile hotspot device. Further still, use of the Li-Ion cylindrical battery pack(s) allows for unique, interactive information display, as well as evolutionary mobile hotspot device design and form factors.

Abstract: A wireless modem card comprises a host interface configured to interface the wireless modem card with a host device of a first type. The wireless modem card can be configured to detect whether a modem interface device configured to interface the wireless modem card with a host device of a second type is communicatively coupled with the host interface. The wireless modem card is also configured to reconfigure the host interface in response to the presence of the modem interface device to enable the wireless modem card to interface with the host device of the second type through the modem interface device.

Abstract: An integrated circuit package includes a balanced-part structure. The condition of thermal stress of chips connected on a substrate decides the amount, locations, weights, and the material of at least a balanced-part fastened on a substrate. The balanced-part is fastened on the substrate to balance stress distribution before an adhering heat sinks process and a packaging molding compound process. The balanced-parts also decreases thermal stress affection and avoid warpage defects of the integrated circuit packages.