Undergraduate
Academic
Internships
Automatic Search for Robust Hybrid Deep Learning Models
Undergraduate Thesis Project, Kyushu University, Fukuoka, Japan
Guide: Dr. Danilo Vasconcellos Vargas, and Prof. Junichi Murata
Sponsor: JST-ACT I
January 17, 2019 – June 17, 2019
Evolutionary Reinforcement Learning
Research, Kyushu University, Fukuoka, Japan
Guide: Dr. Danilo Vasconcellos Vargas, and Prof. Kouichi Sakurai
May 16, 2018 – July 20, 2018
Worked on various reinforcement learning algorithms used for training an agent to play Atari games. Learned about DQN, DDQN, Duelling-DQN, DDPG, A3C, A2C, ACKTR, World Models and other Actor-Critic algorithms. Analysed the algorithms to understand the differences in learning and time taken by an agent to learn. Incorporated various evolutionary strategies such as Differential Evolution (DE) and CMA-ES. Incorporated different model architectures such as DC-GAN BE-GAN, Cycle GAN and Siamese network on the existing RL algorithms to see the performance boost.
Affective Interaction & Sensitive Classroom
Research, Indian Institute of Technology -Guwahati (IIT-G), Guwahati, Assam, India
Guide: Dr. Samit Bhattacharya
December 7, 2017 – January 12, 2018
Worked on a module of the vedinkakSa mobile application, a sensitive ICT-based system to enhance teaching and learning in a large classroom. Contributed in conducting nation-wide survey titled, “Smartphone-Usage by Indian Students” to understand the general usage pattern and activities done by a student in and outside of a classroom. Developed a mobile application to log available system and sensor data in a CSV file for further processing. Developed a hierarchical hybrid classifier of 1D Convolution layer and LSTM to classify an activity by a student in a classroom as study related or non-study related activity. Designed a controlled experiment consisting of 3 study related tasks and 5 non-study related tasks to gather data for training the classifier.
(Link to the related article)Finding Motifs in Time Series Data
Research, National Chiao Tung University(NCTU), Hsinchu, Taiwan (ROC)
Guide: Prof. Wen-Chih Peng
May 22 – July 21 , 2017
The internship is focused on finding motifs in a time series using PM2.5 sensors data from sources like LASS, LASS4U, Indie, ProbeCube, WEBDUINO, AirBox, AsusAirBox, MAPS, EPA, and CWB. Analysing the two algorithms of detecting motifs using Matrix Profile in one and discretization and dimensionality reduction in other. Applied Co-relational analysis between the two motifs to find the relationship between the motifs a novel idea.
"What you believe is all there is for you. Believe more."
– Akiroq Brost
Last Updated: January 2019