Home » Uncategorized » GUIDE-007: semantic web

GUIDE-007: semantic web

SymSys Symbolic Systems

Symbolic Systems - Stanford

Courses

SYMSYS 100. Minds and Machines. 4 Units.

An overview of the interdisciplinary study of cognition, information, communication, and language, with an emphasis on foundational issues: What are minds? What is computation? What are rationality and intelligence? Can we predict human behavior? Can computers be truly intelligent? How do people and technology interact, and how might they do so in the future? Lectures focus on how the methods of philosophy, mathematics, empirical research, and computational modeling are used to study minds and machines. Undergraduates considering a major in symbolic systems should take this course as early as possible in their program of study.
Same as: LINGUIST 144, PHIL 99, PSYCH 35.

SYMSYS 122. Artificial Intelligence: Philosophy, Ethics, & Impact. 3-4 Units.

Recent advances in computing may place us at the threshold of a unique turning point in human history. Soon we are likely to entrust management of our environment, economy, security, infrastructure, food production, healthcare, and to a large degree even our personal activities, to artificially intelligent computer systems. The prospect of “turning over the keys” to increasingly autonomous systems raises many complex and troubling questions. How will society respond as versatile robots and machine-learning systems displace an ever-expanding spectrum of blue- and white-collar workers? Will the benefits of this technological revolution be broadly distributed or accrue to a lucky few? How can we ensure that these systems respect our ethical principles when they make decisions at speeds and for rationales that exceed our ability to comprehend? What, if any, legal rights and responsibilities should we grant them? And should we regard them merely as sophisticated tools or as a newly emerging form of life? The goal of this course is to equip students with the intellectual tools, ethical foundation, and psychological framework to successfully navigate the coming age of intelligent machines.
Same as: CS 122.

SYMSYS 130. Research Methods in the Cognitive and Information Sciences. 3 Units.

Understanding the different methodological approaches used in disciplines that study cognition and information. Emphasis is on philosophical/analytical, formal/mathematical, empirical, and computational thinking styles, with some attention to other methods as well. What assumptions underlie these methods? How can they be combined? How do practitioners of each discipline think differently about problems, and what are the challenges involved in studying or working across them?.

SYMSYS 150. CRYPTOCURRENCIES SEMINAR. 2 Units.

A weekly seminar allowing students the opportunity to discuss and explore cryptocurrencies from a variety of domains and view points:nn1) Explore the history of fiat currencies, both economically and philosophically. How does Bitcoin mesh in here? What are advantages and disadvantages compared to traditional fiat currencies? (~2 weeks) n2) Contextualize and juxtapose decentralized currencies with respect to TCP/IP, Napster, and other relevant decentralized and cloud protocols. (~2 weeks)n3) Work through and understand Satoshi¿s initial protocol and proof-of-work mining system. What problem did she solve? How? Why was it important? How can we prove it mathematically? What are significant game theoretic and cryptographic weaknesses? What do alternative cryptocurrencies look like? Is there a `best¿ alternative? (~3 weeks)n4) What does ¿Bitcoin as a protocol¿ mean? What can be built on top of it? What¿s being built around it? What does regulation look like? What are hypotheses for the future of digital currencies? How do we explain investor confidence, given regulatory hesitation? (~3 weeks).

SYMSYS 170. Decision Behavior: Theory and Evidence. 3-4 Units.

Introduction to the study of judgment and decision making, relating theory and evidence from disciplines such as psychology, economics, statistics, neuroscience, and philosophy. The development and critique of Homo economicus as a model of human behavior, and more recent theories based on empirical findings. Recommended: background in formal reasoning.
Same as: SYMSYS 270.

SYMSYS 190. Senior Honors Tutorial. 1-5 Unit.

Under the supervision of their faculty honors adviser, students work on their senior honors project. May be repeated for credit.

SYMSYS 191. Senior Honors Seminar. 1 Unit.

Recommended for seniors doing an honors project. Under the leadership of the Symbolic Systems program coordinator, students discuss, and present their honors project.

SYMSYS 196. Independent Study. 1-15 Unit.

Independent work under the supervision of a faculty member. Can be repeated for credit.

SYMSYS 200. Symbolic Systems in Practice. 2-3 Units.

Applying a Symbolic Systems education at Stanford and outside. The basics of research and practice. Students develop and present a project, and investigate different career paths, including academic, industrial, professional, and public service, through interviews with alumni.

SYMSYS 201. ICT, Society, and Democracy. 3 Units.

The impact of information and communication technologies on social and political life. Interdisciplinary. Classic and contemporary readings focusing on topics such as social networks, virtual versus face-to-face communication, the public sphere, voting technology, and collaborative production.

SYMSYS 204. Philosophy of Linguistics. 4 Units.

Philosophical issues raised by contemporary work in linguistics. Topics include: the subject matter of linguistics (especially internalism vs. externalism), methodology and data (especially the role of quantitative methods and the reliance on intuitions), the relationship between language and thought (varieties of Whorfianism and anti-Whorfianism), nativist arguments about language acquisition, and language evolution.
Same as: LINGUIST 204, PHIL 369.

SYMSYS 206. Philosophy of Neuroscience. 4 Units.

Can problems of mind be solved by understanding the brain, or models of the brain? The views of philosophers and neuroscientists who believe so, and others who are skeptical of neurophilosophical approaches to the mind. Historical and recent literature in philosophy and neuroscience. Topics may include perception, memory, neural accounts of consciousness, neurophenomenology, neuroscience and physics, computational models, and eliminativism.
Same as: PHIL 167D, PHIL 267D.

SYMSYS 209. Battles Over Bits. 3 Units.

The changing nature of information in the Internet age and its relationship to human behavior. Philosophical assumptions underlying practices such as open source software development, file sharing, common carriage, and community wireless networks, contrasted with arguments for protecting private and commercial interests such as software patents, copy protection, copyright infringement lawsuits, and regulatory barriers. Theory and evidence from disciplines including psychology, economics, computer science, law, and political science. Prerequisite: PSYCH 40, 55, 70, or SYMBSYS 202.

SYMSYS 210. Learning Facial Emotions: Art and Psychology. 3 Units.

Artistic and psychological learning approaches for emotion recognition from facial expressions. The advantages of learning by image-based microexpressions, subtle expressions, macro expressions, art drawing and actor mimicry when there are cognitive deficits due to conditions such as autism. Comparative analysis uses brain studies, learning theory, and human-computer interaction. Studio component conveys the artistic and psychological approaches. Prerequisites: PSYCH 1, SYMSYS 100 or consent of instructor. Go to http://www.stanford.edu/~dwilkins/Symsys210Enroll.doc to sign up for a Permission Number.

SYMSYS 211. Learning Facial Emotions: Art, Psychology, Human-Computer Interaction. 3 Units.

Learning to recognize facial emotions by drawing a live model versus the psychology method of using classified images of subtle and micro expressions. Dimensions of analysis include cognitive modeling and neuroscience. The design of human-computer interaction systems for people with cognitive deficits such as autism and Aspergers, which integrate the art and psychology approaches using methods such as robot heads, avatars, and facial recognition software. Prerequisites: PSYCH 1 or consent of instructor.

SYMSYS 245. Cognition in Interaction Design. 3 Units.

Note: Same course as 145 which is no longer active. Interactive systems from the standpoint of human cognition. Topics include skill acquisition, complex learning, reasoning, language, perception, methods in usability testing, special computational techniques such as intelligent and adaptive interfaces, and design for people with cognitive disabilities. Students conduct analyses of real world problems of their own choosing and redesign/analyze a project of an interactive system. Limited enrollment seminar taught in two sections of approximatly ten students each. Admission to the course is by application to the instructor, with preference given to Symbolic Systems students of advanced standing. Recommended: a course in cognitive psychology or cognitive anthropology.”.

SYMSYS 255. Building Digital History: Social Movements and Protest at Stanford. 3-5 Units.

A project-based course focused on developing a collaborative history website based on oral and archival history research. Thematic focus is the history of student activism at Stanford. How have political activities such as demonstrations, assemblies, educational events, and nonviolent civil disobedience been organized on campus, and how have they affected Stanford? What lessons can be drawn from the past for students interested in social change? Students will choose historical periods and/or specific social movements for research. Course will feature guest appearances by representatives from a range of social movements at Stanford the past fifty years, and the building of an online repository and community for the collaborative representation and discussion of history.

SYMSYS 255A. Building Digital History: Social Movements and Protest at Stanford. 1 Unit.

Lectures-only version of SYMSYS 255.

SYMSYS 270. Decision Behavior: Theory and Evidence. 3-4 Units.

Introduction to the study of judgment and decision making, relating theory and evidence from disciplines such as psychology, economics, statistics, neuroscience, and philosophy. The development and critique of Homo economicus as a model of human behavior, and more recent theories based on empirical findings. Recommended: background in formal reasoning.
Same as: SYMSYS 170.

SYMSYS 280. Symbolic Systems Research Seminar. 1 Unit.

A mixture of public lectures of interest to Symbolic Systems students (the Symbolic Systems Forum) and student-led meetings to discuss research in Symbolic Systems. Can be repeated for credit. Open to both undergraduates and Master’s students.

SYMSYS 290. Master’s Degree Project. 1-15 Unit.

SYMSYS 291. Master’s Program Seminar. 1 Unit.

Enrollment limited to students in the Symbolic Systems M.S. degree program. May be repeated for credit.

SYMSYS 296. Independent Study. 1-15 Unit.

Independent work under the supervision of a faculty member. Can be repeated for credit.

SYMSYS 298. Peer Advising in Symbolic Systems: Practicum. 1 Unit.

Optional for students selected as Undergraduate Advising Fellows in the Symbolic Systems Program. AFs work with program administrators to assist undergraduates in the Symbolic Systems major or minor, in course selection, degree planning, and relating the curriculum to a career or life plan, through advising and events. Meeting with all AFs for an hour once per week under the direction of the Associate Director. Requires a short reflective paper at the end of the quarter on what the AF has learned about advising students in the program. Repeatable for credit. May not be taken by students who receive monetary compensation for their work as an AF.

SYMSYS 299. Curricular Practical Training. 1 Unit.

Students obtain employment in a relevant research or industrial activity to enhance their professional experience consistent with their degree programs. Meets the requirements for curricular practical training for students on F-1 visas. Students submit a concise report detailing work activities, problems worked on, and key results. May be repeated for credit. Prerequisite: qualified offer of employment and consent of advisor.

  1. Submission to the Symbolic Systems Program office and approval of the following pre-project research documents:
    1. Project Area Statement, endorsed with a commitment from a student’s prospective project adviser no later than May 1 of the academic year prior to the expected graduation year; and
    2. Qualifying Research Paper due no later than the end of the Summer Quarter prior to the expected graduation year.
  2. Completion of a coherent plan of study, to be approved by the Graduate Studies Director in consultation with the student’s adviser and designed to support a student’s project. An initial plan of study should be delineated on the Program Proposal Form prior to the end of the student’s first quarter of study, to be modified at the time of the Project Area Statement with the approval of a student’s adviser and the Graduate Studies Director. The final version of the Program Proposal, which should specify all the courses the student has taken and proposes as fulfillment of the unit requirements for the degree, is due by the end of Finals Week in the quarter prior to the student’s expected graduation quarter (i.e. end of Winter Quarter for a student graduating in the Spring). The plan of study must include courses more advanced than the Symbolic Systems undergraduate core in four main skill areas: formal, empirical, computational, and philosophical; and in at least three of the following departments: Computer Science, Linguistics, Philosophy, and Psychology. More advanced courses in each of the skill areas are defined as follows:

a) formal: a course in logic and computational theory beyond the level of PHIL 151 First-Order Logic. The courses below have been approved. Other courses may be approved if appropriate.

b) empirical: a course drawing on experimental or observational data or methods, beyond the level of Psych 55, Ling 120, or Ling 130A. The courses below are examples of those that have been approved. Other courses may be approved if appropriate.

c) computational: a course involving programming beyond the level of CS 107. The courses below have been approved. Other courses may be approved if appropriate.

  • CS 108 Object-Oriented Systems Design
  • CS 110 Principles of Computer Systems
  • CS 124 From Languages to Information
  • CS 142 Web Applications
  • CS 143 Compilers
  • CS 148 Introduction to Computer Graphics and Imaging
  • CS 193R
  • CS 193S
  • CS 221 Artificial Intelligence: Principles and Techniques
  • CS 224N Natural Language Processing
  • CS 224W Social and Information Networks
  • CS 249A Object-Oriented Programming from a Modeling and Simulation Perspective

d) philosophical: a course in the area of Philosophy of Mind/Language/Science/’Epistemology or Metaphysics at the 200 level or above, certified by the instructor as worthy of graduate credit. The courses below are examples of those that have been approved. Other courses may be approved if appropriate.

3. Completion of three quarters of SYMSYS 291 Master’s Program Seminar.

4. Completion of a substantial project appropriate to the program plan, represented by the M.S. Thesis, the last of the the M.S research documents. The project normally takes three quarters, and work on the project may account for up to 15 units of a student’s program. The thesis must be read and approved for the master’s degree in Symbolic Systems by two qualified readers approved by the program, at least one of whom must be a member of the academic council. A copy of the thesis must be submitted (in both print and electronic forms) to the Associate Director of Symbolic Systems, with the print version including the signatures of each reader indicating approval of the thesis for the degree of Master of Science, no later than 5 pm on the last day of finals week in the quarter of a student’s graduation.

AI Artificial Intelligence Concentration [br-AI-n]

For this concentration, students must take CS 221 to satisfy the core AI requirement. In addition, they must complete a total of six courses from the following list, including at least three of the courses marked in *boldface with surrounding asterisks* (drawn from at least two areas):

    1. Knowledge representation and reasoning: Logic and Automated Reasoning (CS 157); *Rational Agency and Intelligent Interaction (CS 222/Phil 358)*; *Reasoning Methods in AI (CS 227)*; *Structured Probabilistic Models: Principles and Techniques (CS 228)*; Formal Methods for Reactive Systems (CS 256); Modal Logic (Phil 154).
    2. Natural language processing: *Natural-Language Processing (CS 224N/Linguist 280)* or From Languages to Information (CS 124/Linguist 180) (but not both); *Speech Recognition and Synthesis (CS 224S/Linguist 285)*; *Natural Language Understanding (CS 224U/Linguist 188/288)*.
    3. Learning: *Machine Learning (CS 229)*; Approximate Dynamic Programming (MS&E 339); Modern Applied Statistics: Learning (Stat 315A); Modern Applied Statistics: Data Mining (Stat 315B).
    4. Robotics and vision: *Introduction to Robotics (CS 223A)*; *Introduction to Computer Vision (CS 223B)*; Experimental Robotics (CS 225A); Robot Programming Laboratory (CS 225B); *Statistical Techniques in Robotics (CS 226)*, *Motion Planning (CS 326A)*.
    5. Additional topics: *Multi-agent Systems (CS 224M)*; General Game Playing (CS 227B); Topics in Artificial Intelligence (CS 329) [with approval of advisor]; Introduction to Biomedical Informatics: Fundamental Methods (CS 270/Biomedin 210); Introduction to Biomedical Informatics: Biomedical Systems Engineering (CS 271/Biomedin 211); Representations and Algorithms for Computational Molecular Biology (CS 274/BioE 214/Biomedin 214/Gene 214); Phenomenological Foundations of Cognition, Language, and Computation (CS 378).
    6. Mathematical foundations: Game Theory and Economic Applications (Econ 160); Introduction to Linear Dynamic Systems (EE 263); Convex Optimization (EE 364A/B); Information Theory (EE 376A/B); Computability and Logic (Phil 152) [if not taken for the core]; Stochastic Decision Models (MS&E 251); Introduction to Control Design Techniques (Engr 205); Analysis and Control of Nonlinear Systems (Engr 209A); Linear Algebra and Matrix Theory (Math 113); Mathematical Methods for Robotics, Vision, and Graphics (CS205A); Introduction to Automata and Complexity Theory (CS154) [if not taken for the core].

THE UNCERTAIN LIFE OF THE QUANTUM MONAD

BY JAMES MORRIS

Abstract: The Uncertain Life of the Quantum Monad by James Morris.
The human ability to create art poses a problem for artificial intelligence. Can computers create art? The study analyses the creative process, contrasts this to computerized systems, then argues that no current computer can replicate this process. Discussing current artificial intelligence research, the study questions whether future theories could derive more creative computing, but concludes that our lack of understanding of ourselves will always leave something indefinable.

james-morris-uncertain-life-quantum-monads

If this doesn’t terrify you…

Google’s computers OUTWIT their humans

‘Deep learning’ clusters crack coding problems their top engineers can’t

Analysis Google no longer understands how its “deep learning” decision-making computer systems have made themselves so good at recognizing things in photos.

This means the internet giant may need fewer experts in future as it can instead rely on its semi-autonomous, semi-smart machines to solve problems all on their own.

The claims were made at the Machine Learning Conference in San Francisco on Friday by Google software engineer Quoc V. Le in a talk in which he outlined some of the ways the content-slurper is putting “deep learning” systems to work. (You find out more about machine learning, a computer science research topic,here [PDF].)

“Deep learning” involves large clusters of computers ingesting and automatically classifying data, such as things in pictures. Google uses the technology for services such as Android’s voice-controlled search, image recognition, and Google translate.

The ad-slinger’s deep learning experiments caused a stir in June 2012 when a front-page New York Times article revealed that after Google fed its “DistBelief” technology with millions of YouTube videos, the software had learned to recognize the key features of cats.

A feline detector may sound trivial, but it’s the sort of digital brain-power needed to identify house numbers for Street View photos, individual faces on websites, or, say,<SKYNET DISCLAIMER> if Google ever needs to identify rebel human forces creeping through the smoking ruins of a bombed-out Silicon Valley </SKYNET DISCLAIMER>.

Google’s deep-learning tech works in a hierarchical way, so the bottom-most layer of the neural network can detect changes in color in an image’s pixels, and then the layer above may be able to use that to recognize certain types of edges. After adding successive analysis layers, different branches of the system can develop detection methods for faces, rocking chairs, computers, and so on.

What stunned Quoc V. Le is that the software has learned to pick out features in things like paper shredders that people can’t easily spot – you’ve seen one shredder, you’ve seen them all, practically. But not so for Google’s monster.

Learning “how to engineer features to recognize that that’s a shredder – that’s very complicated,” he explained. “I spent a lot of thoughts on it and couldn’t do it.”

It started with a GIF: Image recognition paves way for greater things

Many of Quoc’s pals had trouble identifying paper shredders when he showed them pictures of the machines, he said. The computer system has a greater success rate, and he isn’t quite sure how he could write program to do this.

At this point in the presentation another Googler who was sitting next to our humbleEl Reg hack burst out laughing, gasping: “Wow.”

“We had to rely on data to engineer the features for us, rather than engineer the features ourselves,” Quoc explained.

This means that for some things, Google researchers can no longer explain exactly how the system has learned to spot certain objects, because the programming appears to think independently from its creators, and its complex cognitive processes are inscrutable. This “thinking” is within an extremely narrow remit, but it is demonstrably effective and independently verifiable.

Google doesn’t expect its deep-learning systems to ever evolve into a full-blown emergent artificial intelligence, though. “[AI] just happens on its own? I’m too practical – we have to make it happen,” the company’s research chief Alfred Spector told us earlier this year.

Google’s AI chief Peter Norvig believes the kinds of statistical data-heavy models used by Google represent the world’s best hope to crack tough problems such as reliable speech recognition and understanding – a contentious opinion, and one that clashes with Noam Chomsky’s view.

Deep learning is attractive to Google because it can solve problems the company’s own researchers can’t, and it can let the company hire fewer inefficient meatsackshuman experts. And Google is known for hiring the best of the best.

By ceding advanced capabilities to its machines, Google can save on human headcount, better grow its systems to deal with a data deluge, and develop capabilities that have – so far – befuddled engineers.

The advertising giant has pioneered a similar approach of delegating certain decisions and decision-making selection systems with its Borg and Omega cluster managers, which seem to behave like “living things” in how they allocate workloads.

Given Google’s ambition to “organize the world’s information”, the fewer people it needs to employ, the better. By developing these “deep learning” systems Google needs to employ fewer human experts, Quoc, said.

“Machine learning can be difficult because it turns out that even though in theory you could use logistic regression and so on, but in practice what happens is we spend a lot of time on data processing inventing features and so on. For every single problem we have to hire domain experts,” he added.

“We want to move beyond that … there are certainly problems we can’t engineer features of and we want machines to do that.”

By working hard to give its machines greater capabilities, and local, limited intelligence, Google can crack classification problems that its human experts can’t solve. Skynet? No. Rise of the savant-like machines? Yes. But for now the relationship is, thankfully, cooperative.

By Jack Clark

RESEARCH AREAS: AI / Information SystemsComputer Systems / Theory

Intelligent Machines

ELIZA – A Computer Program for the Study of Natural Language Communication between Man and Machine

BRAIN and SPACE / Behavioral Neuroscience / Clinical Psychology

Data Mining | Data Structures and Algorithms (Website / Lecture-1 vid) | Big Data Analytics (Lecture-1 vid)

Brain Teasers

IQ
Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: