The projects that we have developed for UC Berkeley’s introductory artificial intelligence (AI) course teach foundational concepts using the classic video game Pac-Man. There are four project topics: state-space search, multi-agent search, probabilistic inference, and reinforcement learning. Each project requires students to implement general-purpose AI algorithms and then to inject domain knowledge about the Pac- Man environment using search heuristics, evaluation functions, and feature functions. We have found that the Pac-Man theme adds consistency to the course, as well as tapping in to students’ excitement about video games.

Reasoning by the usage of mental images has been the subject of much debate in Cognitive Science, especially among the schools of depictive and descriptive imagistic representations. Whether or not reasoning with mental images involves a mechanism or a process different from language based reasoning is an important question. This paper proposes that any theory which aims for a cohesive whole needs to be constrained by neurophysiological data and such data can be obtained by the Clock Drawing Test. The Clock Drawing Test (CDT) is a screening tool for cognitive impairment and can be used as a tool to test resilience of certain factors of visual spatial representations. Thus, it can help to form an empirical case for which factors are prone to debility and which factors are not during the onset and progress of cognitive impairment from a mental representation point of view. This paper presents 50 CDT tests done on patients with cognitive impairment and analyses the results which support the case for a depictive rather than a descriptive theory for imagistic representations. Lastly, this paper proposes that there is some evidence for a more dynamic and distributed nature of representation in the observations which question the above dichotomy and can be partly explained by certain aspects of the connectionist school of thought.

The architectural approach to AI focuses on the fixed structure underlying intelligence. Applying it to statistical relational AI should further stimulate the application of statistical relational techniques across AI, while focusing research on their commonalities, (in)compatibilities and integration. It could also yield new architectures that are simpler yet more comprehensive than today’s best.

It is still unclear what metacognition is. Two main theories about metacognition are reviewed, each of which claims to provide a better explanation of the phenomenon, while discrediting the other theory as inappropriate. My claim is that in order to do justice to the complex phenomenon of metacognition, we must distinguish two levels of this capacity. It can be shown that each of these theories has been trying to explain only one of the two levels and that, consequently, the conflict between them can be dissolved. Finally, I characterize each level and explain some of their interactions.

We present an AI called Visuo that guesses quantitative visuospatial magnitudes (e.g., heights, lengths) given adjective-noun pairs as input (e.g., “big hat”). It uses a database of tagged images as memory and infers unexperienced magnitudes by analogy with semantically-related concepts in memory. We show that transferring width-height ratios from a semantically-related concept yields significantly lower error rates than using dissimilar concepts when predicting the width-height ratios of novel inputs.

Increasingly, we need to computationally understand real-time streams of information in places such as news feeds, speech streams, and social networks. We present Streaming AnalogySpace, an efficient technique that discovers correlations in and makes predictions about sparse natural-language data that arrives in a real-time stream. AnalogySpace is a noise-resistant PCA-based inference technique designed for use with collaboratively collected common sense knowledge and semantic networks. Streaming AnalogySpace advances this work by computing it incrementally using CCIPCA, and keeping a dense cache of recently-used features to efficiently represent a sparse and open domain. We show that Streaming AnalogySpace converges to the results of standard AnalogySpace, and verify this by evaluating its accuracy empirically on common-sense predictions against standard AnalogySpace.

This paper investigates how readily an unsolvable constraint satisfaction problem can be reformulated so that it becomes solvable. We investigate small changes in the definitions of the problemís constraints, changes that alter neither the structure of its constraint graph nor the tightness of its constraints. Our results show that structured and unstructured problems respond differently to such changes, as do easy and difficult problems taken from the same problem class. Several plausible explanations for this behavior are discussed.

Activity recognition plays an important role in many areas such as smart environments by offering unprecedented opportunities for assisted living, automation, security and energy efficiency. It’s also an essential component for planning and plan recognition in smart environments. One challenge of activity recognition is the need for collecting and annotating huge amounts of data for each new physical setting in order to be able to carry out the conventional activity discovery and recognition algorithms. This extensive initial phase of data collection and annotation results in a prolonged installation process and excessive time investment for each new space. In this paper we propose a new method of transferring learned knowledge of activities to a new physical space in order to leverage the learning process in the new environment. Our method called ”Home to Home Transfer Learning” (HHTL) is based on using a semi EM framework and modeling activities using structural, temporal and spatial features. This method allows us to avoid the tedious task of collecting and labeling huge amounts of data in the target space, and allows for a more accelerated and more scalable deployment cycle in the real world. It also allows us to exploit the insights learned in previous spaces. To validate our algorithms, we use the data collected in several smart apartments with different physical layouts.

We present a game-theoretic model that formalizes core ideas of conversational grounding theory. This game-theoretic model is based on the concept of signaling games, originally proposed as a model of linguistic convention. We extend signaling games with an observation model, which allows for the possibility that the actions a dialog participant takes may only be partially observable to others. We then apply this model to the domain of referential communication tasks, a type of task commonly used in psycholinguistic experiments.

The Model AI Assignments session seeks to gather and disseminate the best assignment designs of the Artificial Intelligence (AI) Education community. Recognizing that assignments form the core of student learning experience, we here present abstracts of eight AI assignments that are easily adoptable, playfully engaging, and flexible for a variety of instructor needs.