Tai-Ning Yang

Bitboard knowledge base system and elegant search architectures for Connect6

Efficiency is critical for game programs. This paper improves the search efficiency of Connect6 program by encoding connection patterns and computing the inherent information in advance. Such information is saved in a bitboard knowledge base system, where special bitwise operations are designed. This paper also proposes efficient methods of generating threat moves and the Multistage Proof Number Search. The methods reduce the time complexity of generating threat moves. The search improves the search performance by developing candidate moves in stages according to their importance. In brief, this paper proposes an efficient knowledge base system and elegant search architectures for Connect6. It is expected that the proposed methods can be applied to all kinds of Connect-k games.

Pattern Matching in Go Game Records

Many professional Go games, Go life-and-death problems and others are saved as digital game records by SGF (smart game format). Valuable information hides in these records. This article presents a novel Go game record information retrieval system. In this system, the most difficult part is Go pattern matching in Go game records. In this article, a Go pattern matching algorithm is given to find game records that contain a desired query pattern. Then, a proposed index structure for a Go record database integrates methods of information retrieval and domain knowledge of Go. This index can increase the speed of pattern matching in the game database.

Selective-tracing waveform relaxation algorithm for incremental circuit simulation

This paper discusses the large-scale circuit simulation and incremental circuit simulation, both which are based on selective-tracing waveform relaxation (STWR) algorithm. Utilizing the concept of exactness in simulating subcircuits, STWR usually spends fewer subcircuit calculations than traditional algorithms, such as WR and ITA. But STWR suffers from robustness problems. Techniques are proposed to reinforce characteristics of STWR. This paper also describes the incremental circuit simulation that is based on STWR. This new method manages the influence of design changes more exactly than previous works, and hence exhibits better efficiency. We have implemented all proposed methods and performed experiments. Simulation results justify better characteristics of proposed methods.

A Simple and Rapid Lights-up Solver

This study proposes a solver that mix pattern matching method with elimination search to solve Lights-up puzzles. First, the pattern matching method is used to solve some parts of puzzle in a very short time. Then, the elimination search method is applied to solve the part in which the former method cannot be applied. The solver is authenticated by solving problems taken from Internet and compared with methods in other studies. The proposed approach is fast and efficient.

M-eval: A Multivariate Evaluation Function for Opening Positions in Computer Go

Recently, computer go has experienced great advance with the introduction of Monte-Carlo Tree Search (MCTS). Although MCTS programs are overall stronger than previous programs, their strength manifests mostly as the game advances. Strong human players applying established opening principles overtake current MCTS programs in the early moves of nonhandicap 19x19 games. In this paper, the authors propose a method, M-eval, for implementing these opening principles in a program. M-eval is a multivariate evaluation function that provides a priori information for MCTS-based go programs. An initial vector made of k positive qualities both players wish to maximize is computed on a board position for each player. The initial vector is dimensionally reduced using non-negative matrix factorization to minimize informational loss. This reduction is done using the multivariate structure learned offline from a set of over 110,000 board positions from a database of master human games. The resulting R2 encoding is converted by regression to a real value used as a priori information in MCTS. Games were played using the MCTS GoKnot/QYZ framework with M-eval enabled against the same program without M-eval. Results show valuable improvement both in territorial gain and number of wins.

Direct Minutiae Matching in Gray-Level Fingerprint Images

Most researchers applied the same procedure in both fingerprint enrollment and fingerprint verification. In general, many overall pixel-level computations are employed in the procedure of fingerprint enrollment to obtain clear minutiae. Because the minutiae-matching is performed only once for one-to-one verification, we propose a different procedure to directly check the positions of minutiae in gray-level fingerprint images. Therefore, this approach avoids some time-consuming overall pixel-level computations and achieves a fast minutiae matching

The Backward-traversing Relaxation Algorithm for Circuit Simulation

This paper proposes a new relaxation-based circuit simulation algorithm that is more robust and efficient than traditional methods such as waveform relaxation (WR) and iterated timing analysis (ITA). The new method employs a brand new strategy to simulate: it simulates by performing depth-first search in the signal flow graph of simulated circuits. The new method flexibly schedules subcircuits for calculating according to converging situations of subcircuits, so it can achieve robustness as well as efficiency in dealing with various types of circuits. A circuit simulation program based on the proposed method has been implemented, and various circuits have been tested to justify its performance

Apply principal Gabor basis functions to extract discontinuous points from low-resolution binarized fingerprint images

In fingerprint images, the discontinuous points include cores, deltas, minutiae, and etc. Generally, cores and deltas are applied to fingerprint classification and minutiae are to fingerprint matching. Owing to complicated processes, most researchers developed minutiae extraction algorithm from 500 dpi fingerprint images in the past. Therefore, much time is wasted before minutiae are extracted. Besides, most researchers concentrated on how to use Gabor filters to capture the continuity of ridge structures. On the contrast, we will apply Gabor filters to detect the discontinuity of ridge structures. Therefore we will develop a low-resolution and fewer steps approach to extract minutiae by using Gabor filters in this paper.

Singular Point Detection in Fingerprint Images by a Bank of Discrete Fourier Filters

Singular point detection is a critical process for both fingerprint matching and fingerprint classification. The process of singular points detection must be fast and robust; otherwise, the performance of the whole fingerprint recognition system would be influenced heavily. In this paper, we will combine the advantages of Poincaré index and fast Fourier transform to develop a fast and robust approach for singular point detection and to avoid the time-consuming problem of Gabor transform-based approach.

Fuzzy objective functions for robust pattern recognition

In this paper, we consider the issue of fuzzy objective functions when outliers exist. The outlier set is defined as the complement of the data set. Following this concept, a specially designed fuzzy membership weighted objective function is proposed and the corresponding optimal membership is derived. Based on the proposed robust objective functions, algorithms for clustering are implemented. Artificially generated data are used for comparison.