Projects at Michigan State University

BEACON | An NSF Center for the Study of Evolution in Action

Evolving Multi-Layer Markov Network Brains Using Adaptive Complexification

• Developed an adaptive complexification method for MNBs to evolve the necessary complexity level in regards to layers and hidden nodes in addition to their connections in the network to solve boolean logic problems and Super Mario Bros.

Evolving Quick Learners: Novel Initialization Strategies for

Markov Network Brains Abstract:

• Developed ‘‘pre-evolution’’ multi-task pre-training strategies for better initialization of MNBs.

Projects at The Pennsylvania State University

Master's Thesis:

A Clustering Approach to the Bounded Diameter Minimum Spanning Tree Problem Using Ants

Download Thesis(pdf)
Master's Advisor:
Dr. Thang N. Bui at Penn State Harrisburg
Abstract:
The bounded diameter minimum spanning tree problem is the problem of finding a minimum cost spanning tree of a graph such that the number of edges along the longest path in the tree is at most d. This problem is well known to be NP-hard. We present an ant-based algorithm for this problem in which we use two species of ants. The first species is used to discover clusters in the vertex set and then a bounded diameter spanning tree (BDST) is created within each cluster. The second species is then used to connect the cluster BDSTs together building a bounded diameter spanning tree for the whole graph. This tree is then locally optimized yielding a solution to the overall problem. Experimental tests have been conducted on two types of complete graphs, 135 Euclidean and 120 non-Euclidean, totalling 255 graphs. Each Euclidean graph consists of a set of vertices randomly placed in the unit square and the edge cost between two vertices is the Euclidean distance between them. The non-Euclidean graphs are structured such that all of the edge weights in the graph have been randomly selected from 0.01,0.99. For the Euclidean graphs the results show that our algorithm achieves solutions close to the best known for most of the instances. However, on the non-Euclidean graphs our algorithm has obtained new best known solutions for the majority of the graphs and has come very close to the best known in the others.

Yue Lab

A Supervised Learning Approach to the Prediction of Hi-C Data

Download Poster(pdf)
Supervisor: Dr. Feng Yue at Penn State College of Medicine

• Developed a machine learning framework for the prediction of 3-dimensional genome organization data, more specifically high-throughput chromatin conformation capture (Hi-C).

• The framework further allowed cross-tissue/cell prediction helping to gain insights into tissue-specific gene expression while discovering novel biological insights into which genomic/epidenomic features are critical towards chromatin interactions.

Other

Parallel Distributed Genetic Algorithm for Feature Selection

Summary:

• Created an island model distributed genetic algorithm using MPI to perform feature selection.

Micromouse for the IEEE Region 2 Student Activites Conference

• Worked in a team to design, build, and program a robotic mouse to solve the IEEE maze

Software Verification and Security Analysis by Modeling System Specifications

• Creating statecharts, modeling them using PROMELA, and designing safety/liveness properties in Linear Temporal Logic (LTL) to prove correctness using the Spin Model Checker

Voice-to-Braille Translation System

• Worked in a team to design and create a refreshable braille display based on utilizing an Arduino, Android App that communicated via bluetooth to our custom refreshable braille device