If you want to know what motivated me to create TrueShelf (an AI powered adaptive learning platform) and the future of learning, please read my interview on Edsurge.
Wish you all a Very Happy New Year. Here is a list of my 10 favorite open problems for 2014. They belong to several research areas inside discrete mathematics and theoretical computer science. Some of them are baby steps towards resolving much bigger open problems. May this new year shed new light on these open problems.
- 1. Combinatorics : Prove that trees with diameter 6 are graceful. (see earlier posts graceful tree conjecture, graceful diameter-6 trees).
- 2. Optimization : Improve the approximation factor for the undirected graphic TSP. The best known bound is 7/5 by Sebo and Vygen.
- 3. Algorithms : Prove that the tree-width of a planar graph can be computed in polynomial time (or) is NP-complete.
- 4. Fixed-parameter tractability : Treewidth and Pathwidth are known to be fixed-parameter tractable. Are directed treewidth/DAG-width/Kelly-width (generalizations of treewidth) and directed pathwidth (a generalization of pathwidth) fixed-parameter tractable ? This is a very important problem to understand the algorithmic and structural differences between undirected and directed width parameters.
- 5. Space complexity : Is Planar ST-connectvity in logspace ? This is perhaps the most natural special case of the NL vs L problem. Planar ST-connectivity is known to be in . Recently, Imai, Nakagawa, Pavan, Vinodchandran and Watanabe proved that it can be solved simultaneously in polynomial time and approximately O(√n) space.
- 6. Metric embedding : Is the minor-free embedding conjecture true for partial 3-trees (graphs of treewidth 3) ? Minor-free conjecture states that “every minor-free graph can be embedded in with constant distortion. The special case of planar graphs also seems very difficult. I think the special case of partial 3-trees is a very interesting baby step.
- 7. Structural graph theory : Characterize pfaffians of tree-width at most 3 (i.e., partial 3-trees). It is a long-standing open problem to give a nice characterization of pfaffians and design a polynomial time algorithm to decide if an input graph is a pfaffian. The special of partial 3-trees is an interesting baby step.
- 8. Structural graph theory : Prove that every minimal brick has at least four vertices of degree three. Bricks and braces are defined to better understand pfaffians. The characterization of pfaffian braces is known (more generally characterization of bipartite pfaffians is known). To understand pfaffians, it is important to understand the structure of bricks. Norine,Thomas proved that every minimal brick has at least three vertices of degree three and conjectured that every minimal brick has at least cn vertices of degree three.
- 9. Communication Complexity : Improve bounds for the log-rank conjecture. The best known bound is
- 10. Approximation algorithms : Improve the approximation factor for the uniform sparsest cut problem. The best known factor is .
Here are my conjectures for 2014 🙂
- Weak Conjecture : at least one of the above 10 problems will be resolved in 2014.
- Conjecture : at least five of the above 10 problems will be resolved in 2014.
- Strong Conjecture : All of the above 10 problems will be resolved in 2014.
Have fun !!
One year back (on 6/6/12) I announced a beta version of TrueShelf, a social-network for sharing exercises and puzzles especially in mathematics and computer science. After an year of testing and adding new features, now I can say that TrueShelf is out of beta.
TrueShelf turned out to be a very useful website. When students ask me for practice problems (or books) on a particular topic, I simply point them to trueshelf and tell them the tags related to that topic. When I am advising students on research projects, I first tell them to solve all related problems (in the first couple of weeks) to prepare them to read research papers.
Here are the features in TrueShelf 1.0.
- Post an exercise (or) multiple-choice question (or) video (or) notes.
- Solve any multiple-choice question directly on the website.
- Add topic and tags to any post
- Add source or level (high-school/undergraduate/graduate/research).
- Show text-books related to a post
- Show related posts for every post.
- View printable version (or) LaTex version of any post.
- Email / Tweet / share on facebook (or) Google+ any post directly from the post.
- Add any post to your Favorites
- Like (a.k.a upvote) any post.
Feel free to explore TrueShelf, contribute new exercises and let me know if you have any feedback (or) new features you want to see. You can also follow TrueShelf on facebook, twitter and google+. Here is a screenshot highlighting the important features.
During summer 2010 I started reading a book titled Elements of Automata Theory by Jacques Sakarovitch. It took me one year to read the book and submit my review to Bill Gasarch during summer 2011. It will be appearing in SIGACT book review column soon. I am posting my review here for the benefit of everybody.
Book : Elements of Automata Theory by Jacques Sakarovitch
Reviewer : Shiva Kintali
During my undergrad I often found myself captivated by the beauty and depth of automata theory. I wanted to read one book on automata theory and say that I “know” automata theory. Couple of years later I realized that it is silly to expect such a book. The depth and breadth of automata theory cannot be covered by a single book.
My PhD thesis is heavily inspired by automata theory. I had to read several (fifty year old) papers and books related to automata theory to understand several fundamental theorems. Unfortunately, the concepts I wanted to learn are scattered in multiple books and old research papers, most of which are hard to find. When I noticed that Prof. Bill Gasarch is looking for a review of Elements of Automata Theory, I was very excited and volunteered to review it, mainly because I wanted to increase my knowledge about automata theory. Given my background in parsing technologies and research interests in space-bounded computation I wanted to read this book carefully. This book is around 750 pages long and it took me around one year to (approximately) read it. It is very close to my expectations of the one book on automata theory.
First impressions : Most of the books on automata theory start with the properties of regular languages, finite automata, pushdown automata, context-free languages, pumping lemmas, Chomsky hierarchy, decidability and conclude with NP-completeness and the P vs NP problem. This book is about “elements” of automata theory. It focuses only on finite automata over different mathematical structures. It studies pushdown automata only in the context of rational subsets in the free group. Yes, there is 750 pages worth literature studying only finite automata.
This book is aimed at people enthusiastic to know the subject rigorously and not intended as a textbook for automata theory course. It can also be used by advanced researchers as a desk reference. There is no prerequisite to follow this book, except for a reasonable mathematical maturity. It can be used as a self-study text. This book is a direct translation of its french original.
This book is divided into five major chapters. The first three chapters deal with the notions of rationality and recognizability. A family of languages are rationally closed if it is closed under rational operations (union, product and star). A language is reconizable if there exists a finite automata that recognizes it. The fourth and fifth chapters discuss rationality in relations. Chapter 0 acts as an appendix of several definitions of structures such as relations, monoids, semirings, matrices, graphs and concepts such as decidability. Following is a short summary of the five major chapters. There are several deep theorems (for example, Higman’s Theorem) studied in this book. I cannot list all of them here. The chapter summaries in the book have more details.
Chapter 1 essentially deals with the basic definitions and theorems required for any study of automata theory. It starts with the definitions of states, transitions, (deterministic and nondeterministic) automaton, transpose, ambiguity and basic operations such as union, cartesian product, star, quotient of a language. Rational operators, rational languages and rational expressions are defined and the relation between rationality and recognizability is established leading to the proof of Kleene’s theorem. String matching (i.e., finding a word in a text) is studied in detail as an illustrative example. Several theorems related to star height of languages are proved. A fundamental theorem stating that `the language accepted by a two-way automaton is rational’ is proved. The distinction between Moore and Mealy machines is introduced.
Chapter 2 deals with automata over the elements of an arbitrary monoid and the distinction between rational set and recognizable set in this context. This leads to a better understanding of Kleene’s theorem. The notion of morphism of automata is introduced and several properties of morphisms and factorisations are presented. Conway’s construction of universally minimal automaton is explained and the importance of well quasi-orderings is explained in detail. Based on these established concepts, McNaughton’s theorem (which states that the star height of a rational group language is computable) is studied with a new perspective.
Chapter 3 formalizes the notion of “weighted” automata that count the number of computations that make an element be accepted by an automaton, thus generalizing the previously introduced concepts in a new dimension. Languages are generalized to formal series and actions are generalized to representations. The concepts and theorems in this chapter makes the reader appreciate the deep connections of automata theory with several branches of mathematics. I personally enjoyed reading this chapter more than any other chapter in this book.
Chapter 4 builds an understanding of the relations realized by different finite automata in the order they are presented in chapters 1, 2 and 3. The Evaluation Theorem and the Composition Theorem play a central role in understanding this study. The decidability of the equivalence of transducers (with and without weigths) is studied. This chapter concludes with the study of deterministic and synchronous relations.
Chapter 5 studies the functions realized by finite automata. Deciding functionality, sequential functions, uniformisation of rational relations by rational functions, semi-monomial matrix representation, translations of a function and uniformly bounded functions are studied.
There are exercises (with solutions) at the end of every section of every chapter. These exercises are very carefully designed and aid towards better understanding of the corresponding concepts. First time readers are highly encouraged to solve (or at least glance through) these exercises. Every section ends with Notes and References mentioning the corresponding references and a brief historical summary of the chapter.
Overall I found the book very enlightening. It has provided me new perspectives on several theorems that I assumed I understood completely. Most of the concepts in this book are new to me and I had no problems following the concepts and the corresponding theorems. The related exercises made these topics even more fun to learn. It was a joy for me to read this book and I recommend this book for anyone who is interested in automata theory (or more generally complexity theory) and wants to know the fundamental theorems of theory of computing. If you are a complexity theorist, it is worthwhile to look back at the foundations of theory of computing to better appreciate its beauty and history. This book is definitely unique in its approach and the topics chosen. Most of the topics covered are either available in very old papers or not accesible at all. I applaud the author for compiling these topics into a wonderful free-flowing text.
This book is nicely balanced between discussions of concepts and formal proofs. The writing is clear and the topics are organized very well from the most specific to the most general, making it a free-flowing text. On the other hand, it is very dense and requires lots of motivation and patience to read and understand the theorems. The author chose a rigorous way of explaining rationality and recognizability. Sometimes you might end up spending couple of hours to read just two pages. Such is the depth of the topics covered. Beginners might find this book too much to handle. I encourage beginners to read this book after taking an introductory automata theory course. This is definitely a very good reference text for researchers in the field of automata theory.
In terms of being used in a course, I can say that a graduate level course can be designed from a carefully chosen subset of the topics covered in this book. The exercises in the book can be readily used for such a course.
This is an expensive book, which is understandable based on the author’s efforts to cover several fundamental topics (along with exercises) in such a depth. If you think it is expensive, I would definitely suggest that you get one for your library.
I have been teaching (courses related to algorithms and complexity) for the past six years (five years as a PhD student at GeorgiaTech, and the past one year at Princeton). One of the most challenging and interesting part of teaching is creating new exercises to help teach the important concepts in an efficient way. We often need lots of problems to include in homeworks, midterms, final exams and also to create practice problem sets.
We do not get enough time to teach all the concepts in class because the number of hours/week is bounded. I personally like to teach only the main concepts in class and design good problem sets so that students can learn the generalizations or extensions of the concepts by solving problems hands-on. This helps them develop their own intuitions about the concepts.
Whenever I need a new exercise I hardly open a physical textbook. I usually search on internet and find exercises from a course website (or) “extract” an exercise from a research paper. There are hundreds of exercises “hidden” in pdf files across several course homepages. Instructors often spend lots of time designing them. If these exercises can reach all the instructors and students across the world in an efficiently-indexed form, that will help everybody. Instructors will be happy that the exercises they designed are not confined to just one course. Students will have an excellent supply of exercises to hone their problem-solving skills.
During 2008, half-way through my PhD, I started collected the exercises I like in a private blog. At the same time I registered the domain trueshelf.com to make these exercises public. In 2011, towards the end of my PhD, I started using the trueshelf.com domain and made a public blog so that anybody can post an exercise. [ Notice that I did not use the trueshelf.com domain for three years. During these three years I got several offers ranging upto $5000 to sell the domain. So I knew I got the right name 🙂 ] Soon, I realized that wordpress is somewhat “static” in nature and does not have enough “social” features I wanted. A screenshot of the old website is shown below.
The new version of TrueShelf is a social website enabling “crowd-sourcing” of exercises in any area. Here is the new logo, I am excited about 🙂
The goal of TrueShelf is to aid both the instructors and students by presenting quality exercises with tag-based indexing. Read the TrueShelf FAQ for more details. Note that we DO NOT allow users to post solutions. Each user may add his own “private” solution and notes to any exercise. I am planning to add more features soon.
In the long-run, I see TrueShelf becoming a “Youtube for exercises”. Users will be able to create their own playlists of exercises (a.k.a problem sets) and will be recommended relevant exercises. Test-preparation agencies will be able to create their own channels to create sample tests.
Let’s see how TrueShelf evolves.
Today’s post is a continuation of earlier posts (here, here, here, here) on graph isomorphism, treewidth and pathwidth. As mentioned earlier, the best known upper bound for Graph Isomorphism of partial k-trees is LogCFL.
Theorem ([Das, Toran and Wagner’10]) : Graph isomorphism of bounded treewidth graphs is in LogCFL.
One of the bottlenecks of the algorithm of [DTW’10] is computing bounded tree decompositions in logspace. This is recently resolved by an amazing result of Elberfeld, Jakoby and Tantau [EJT’10]. The results in this paper are very powerful. Unfortunately, it is still not clear how to improve the LogCFL upper bound.
Can we improve the upper bound for special cases of partial k-trees ? How about bounded pathwidth graphs ? Again, one bottleneck here is to compute bounded path decompositions in logspace. [EJT’10]’s paper does not address this bottleneck and it is not clear how to extend their algorithm to compute path decompositions.
In joint work with Sinziana Munteanu, we resolved this bottleneck and proved the following theorem. Sinziana is a senior undergraduate student in our department. She is working with me on her senior thesis.
Theorem (Kintali, Munteanu’12) : For all constants , there exists a logspace algorithm that, when given a graph of treewidth , decides whether the pathwidth of is at most , and if so, finds a path decomposition of of width in logspace.
A draft of our results is available here. The above theorem is a logspace counterpart of the corresponding polynomial-time algorithm of [Bodlaender, Kloks’96]. Converting it into a logspace algorithm turned out to be a tedious task with some interesting tricks. Our work motivates the following open problem :
Open problem : What is the complexity of Graph Isomorphism of bounded pathwidth graphs ? Is there a logspace algorithm ?
Stay tuned for more papers related to graph isomorphism, treewidth and pathwidth. I am going through a phase of life, where I have more results than I can type. Is there an app that converts voice to latex ? Is there a journal that accepts hand-written proofs ? 🙂
Every once in a while, I can’t help thinking about “the complexity of graph isomorphism for bounded treewidth graphs“. Today has been one of those days again. See my earlier post to get the context.
Theorem ([Das, Toran and Wagner’10]) : Graph isomorphism of bounded treewidth graphs is in LogCFL.
The proof of the above theorem is as follows
- Graph isomorphism of bounded tree-distance width graphs is in L.
- Given two graphs and their tree decompositions, computing the isomorphism respecting these tree decompositions is reducible to (1).
- Given tree decomposition of only one graph, we can guess the tree decomposition of the other and guess the isomorphism (respecting the tree bags) and verify them using a non-deterministic auxiliary pushdown automata (a.k.a LogCFL).
- Since tree decomposition of a graph can be computed in LogCFL, the above theorem follows.
Given a graph and a decomposition , how fast can we verify that is a valid tree decomposition of ? The upper bound of LogDCFL (the deterministic version of LogCFL) is clear from the above mentioned results. Can this verification be done in logspace ?
The truth is “this is a cute problem“. I need to do something to take my mind off this problem and move on. Easy enough, except I need an idea.