Techniques for computing global feature explanations using adaptive sampling are provided. In one technique, first and second samples from an dataset are identified. A first set of feature importance values (FIVs) is generated based on the first sample and a machine-learned model. A second set of FIVs is generated based on the second sample and the model. If a result of a comparison between the first and second FIV sets does not satisfy criteria, then: (i) an aggregated set is generated based on the last two FIV sets; (ii) a new sample that is double the size of a previous sample is identified from the dataset; (iii) a current FIV set is generated based on the new sample and the model; (iv) determine whether a result of a comparison between the current and aggregated FIV sets satisfies criteria; repeating (i)-(iv) until the result of the last comparison satisfies the criteria.

Consciousness is widely considered to be a mysterious and uniquely human trait, which cannot be achieved artificially. On the contrary, a system and method are disclosed for a computational machine that can recognize itself and other agents in a dynamic environment, in a way that seems quite similar to biological consciousness in humans and animals. The machine comprises an artificial neural network configured to identify correlated temporal patterns and attribute causality and agency. The machine is further configured to construct a virtual reality environment of agents and objects based on sensor inputs, to create a coherent narrative, and to select future actions to pursue goals. Such a machine may have application to enhanced decision-making in autonomous vehicles, robotic agents, and intelligent digital assistants.

In some examples, machine learning based semantic structural hole identification may include mapping each text element of a plurality of text elements of a corpus into an embedding space that includes embeddings that are represented as vectors. A semantic network may be generated based on semantic relatedness between each pair of vectors. A boundary enclosure of the embedding space may be determined, and points to fill the boundary enclosure may be generated. Based on an analysis of voidness for each point within the boundary enclosure, a set of void points and void regions may be identified. Semantic holes may be identified for each void region, and utilized to determine semantic porosity of the corpus. A performance impact may be determined between utilization of the corpus to generate an application by using the text elements without filling the semantic holes and the text elements with the semantic holes filled.

Methods for automatic language detection for handwritten text are performed by systems and devices. Such automatic language detection is performed prior to sending representations of the handwritten text to a language recognition engine. Handwritten inputs including one or more writing strokes are received from an input interface, and are associated with coordinates of the inputs and times that the inputs are made. The handwritten inputs are grouped into words based on the coordinates and times. Writing strokes are normalized, and then the words are individually transformed to generate language vectors, such as through a recurrent neural network. The language vectors are used to determine language probabilities for the handwritten inputs. Based on the language probabilities, the handwritten inputs are provided to a specific language recognition engine to determine the language thereof prior to translation or transcription.

An audio event detection system that subsamples input audio data using a series of recurrent neural networks to create data of a coarser time scale than the audio data. Data frames corresponding to the coarser time scale may then be upsampled to data frames that match the finer time scale of the original audio data frames. The resulting data frames are then scored with a classifier to determine a likelihood that the individual frames correspond to an audio event. Each frame is then weighted by its score and a composite weighted frame is created by summing the weighted frames and dividing by the cumulative score. The composite weighted frame is then scored by the classifier. The resulting score is taken as an overall score indicating a likelihood that the input audio data includes an audio event.

Consciousness is widely considered to be a mysterious and uniquely human trait, which cannot be achieved artificially. On the contrary, a system and method are disclosed for a computational machine that can recognize itself and other agents in a dynamic environment, in a way that seems quite similar to biological consciousness in humans and animals. The machine comprises an artificial neural network configured to identify correlated temporal patterns and attribute causality and agency. The machine is further configured to construct a virtual reality environment of agents and objects based on sensor inputs, to create a coherent narrative, and to select future actions to pursue goals. Such a machine may have application to enhanced decision-making in autonomous vehicles, robotic agents, and intelligent digital assistants.

Examples of artificial intelligence-based reasoning explanation are described. In an example implementation, a knowledge model having a plurality of ontologies and a plurality of inferencing rules is generated. Once the knowledge model is generated, based on a real-world problem, a knowledge model from amongst various knowledge models is selected to be used for resolving a real-world problem. The data procured from the real-world problem is clustered and classified into an ontology of the determined knowledge model. Inferencing rules to be used for deconstructing the real-world problem are identified, and a machine reasoning is generated to provide a hypothesis for the problem and an explanation to accompany the hypothesis.

In various examples there is a knowledge base construction and/or maintenance system for use with a probabilistic knowledge base. The system has a probabilistic generative model comprising a process for generating text or other formatted data from the knowledge base. The system has an inference component configured to generate inference results, by carrying out inference using inference algorithms, run on the probabilistic generative model, in either a forward direction whereby text or other formatted data is generated, or a reverse direction whereby text or other formatted data is observed and at least one unobserved variable of the probabilistic generative model is inferred. The inference component is configured to update the knowledge base using at least some of the inference results.

A system for configuring and using a logical neural network including a graph syntax tree of formulae in a represented knowledgebase connected to each other via nodes representing each proposition. One neuron exists for each logical connective occurring in each formula and, additionally, one neuron for each unique proposition occurring in any formula. All neurons return pairs of values representing upper and lower bounds on truth values of their corresponding subformulae and propositions. Neurons corresponding to logical connectives accept as input the output of neurons corresponding to their operands and have activation functions configured to match the connectives' truth functions. Neurons corresponding to propositions accept as input the output of neurons established as proofs of bounds on the propositions' truth values and have activation functions configured to aggregate the tightest such bounds. Bidirectional inference permits every occurrence of each proposition in each formula to be used as a potential proof.

Systems, methods, and computer-readable non-transitory storage medium for communicating medical information based at least in part on an oral communication between a doctor and a patient is disclosed. In this method and system, doctor and patient's respective contexts is inferred from the oral communication. It is also preferred that diagnostic information and respective contexts of the communications can be also inferred. Then, a desired impact of a recipient to a written communication related to the oral communication is inferred. Once the desired impact is inferred, the system generates output text using an artificial intelligence system, or by accessing a database of a plurality of stock phrases, to have appropriate surface text, and subtext, and optionally appropriate tone. The output text can be selected as a function of inferred diagnostic information, the inferred doctor and recipient's respective contexts, the desired impact, and the stock phrases.