Capital One Services, LLC

Omar Florez Choque

Erik Thomas Mueller

Anish Khazane

Alan Salimov

Zachary Kulis

A system for using hash keys to preserve privacy across multiple tasks is disclosed. The system may provide training batch(es) of input observations each having a customer request and stored task to an encoder, and assign a hash key(s) to each of the stored tasks. The system may provide a new batch of input observations with a new customer request and new task to the encoder. The encoder may generate a new hash key assigned to the new customer request and determine whether any existing hash key corresponds with the new hash key. If so, the system may associate the new batch of input observations with the corresponding hash key and update the corresponding hash key such that it is also configured to provide access to the new batch of input observations. If not, the system may generate a new stored task and assign the new hash key to it.

Systems using hash keys to preserve privacy across multiple tasks

A machine learning (ML) system is provided for integrating common sense into the ML system. In contrast to existing machine learning algorithms that search for statistical correlations between concepts, the ML system is configured to learn the semantic relationships or causality between the concepts. This may be accomplished by training an algorithm or data structure to learn similar vector representations of words present in the same context (e.g., that are present together in the same sentence). The resulting AI/ML structure may be used to guide the generation of a causal graph having predictive capabilities. This causal graph may represent semantic relationships and/or causation between concepts, and hence may be employed to introduce a degree of common sense in the machine learning system.

Methods, mediums, and systems for generating causal inference structure between concepts having predictive capabilities

A system includes one or more memory devices storing instructions, and one or more processors configured to execute the instructions to perform steps of a method. A method can include receiving observations and a corresponding class label, determining a candidate key based on the observations, determining a current memory state of a memory module based on a similarity of stored keys to the candidate key, generating a measurement vector based on the current memory state, concatenating the candidate key and measurement vector to form a state vector, determining, based on the state vector and an action distribution policy, an action of a plurality of actions such that the determined action maximizes an expected reduction in entropy as compared to the remaining actions of the plurality actions, executing the determined action, determining a value of the determined action, and updating, based on the value, the action distribution policy.

Systems providing a learning controller utilizing indexed memory and methods thereto

Disclosed are an apparatus, a system and a non-transitory computer readable medium that implement processing circuitry that receives non-dialog information from a smart device and determines a data type of data in the received non-dialog information. Based on the determined data type, the processing circuitry transforms the received first data using an input from a machine learning algorithm into transformed data. The transformed data is standardized data that is palatable for machine learning algorithms such as those used implemented as chatbots. The standardized transformed data is useful for training multiple different chatbot systems and enables the typically underutilized non-dialog information to be used to as training input to improve context and conversation flow between a chatbot and a user.

Techniques to add smart device information to machine learning for increased context

Memory augmented neural networks may use one or more neural encoders to transform input data into distributed representations and a memory module to store the representations with individual addresses. Memory augmented neural networks allow for few-shot learning capabilities because latent representations are persistent between training examples and gradient-based updates affect only certain memory locations via content-based lookups. When a query vector is not found in memory and the memory is full, existing memories that are positively associated with a particular representation may be identified, redundant memories may be aged, and updated memories may be generated. These updated memories retain relevant information acquired during training and reduce redundancy in the memories stored using the memory module, thereby improving the efficiency of data storage and reducing overfitting of data typically encountered with existing neural networks using memory modules.

Response generation using memory augmented deep neural networks

A method for determining machine learning training parameters is disclosed. The method can include a processor receiving a first input. The processor may receive a first response to the first input, determine a first intent, and identify a first action. The processor can then determine first trainable parameter(s) and determine whether the first trainable parameter(s) is negative or positive. Further, the processor can update a training algorithm based on the first trainable parameter(s). The processor can then receive a second input and determine a second intent for the second input. The processor can also determine a second action for the second intent and transmit the second action to a user. The processor can then determine second trainable parameter(s) and determine whether the second trainable parameter(s) is positive or negative. Finally, the processor can further update the training algorithm based on the second trainable parameter(s).

Systems and methods for determining training parameters for dialog generation

Systems, methods, and articles of manufacture for learning design policies based on user interactions. One example includes determining a first task for an environment, receiving data from a plurality of data sources, determining a first time step associated with the received data, determining a plurality of candidate actions for the determined first time step, computing a respective probability value of each candidate action achieving the first task at the first time step based on a first machine learning (ML) model, determining that a first candidate action has a greater probability value for achieving the first task at the first time step relative to the remaining plurality of candidate actions, determining that the first candidate action has not been implemented in the environment at the first time step, and generating an indication specifying to implement the first candidate action as part of a policy to achieve the first task.

Learning: learning design policies based on interactions

Memory management in memory augmented deep neural networks

Exemplary embodiments relate to techniques for integrating common sense into a machine learning (ML) system. In contrast to existing machine learning algorithms that search for statistical correlations between concepts, exemplary embodiments attempt to learn the semantic relationships or causality between the concepts. This may be accomplished by training an algorithm or data structure to learn similar vector representations of words present in the same context (e.g., that are present together in the same sentence). The resulting AI/ML, structure may be used to guide the generation of a causal graph having predictive capabilities. This causal graph may represent semantic relationships and/or causation between concepts, and hence may be employed to introduce a degree of common sense in the machine learning system.

Growing community of inventors

Oakland, CA, United States of America

Omar Florez Choque