I love learning different facets of Computer Science and explore deeper areas for research and analysis. Areas of my present interest and the areas I would like to deep drive are:
- Artificial Intelligence: AI is one of the areas I want to pursue and learn a lot more about. I believe the potential of this field is limitless and more development and research over the years will surely lead to breakthroughs, which will change the world for the better.
- Cognitive Robotics: I want to work towards pushing the field of robotics beyond automation to cognition, which would allow machines to have human-like perception and intelligence. This would extensively broaden the scope of usage of robotics in various fields.
- Algorithms: They are the fundamentals behind computer science; it is the Computer Science Engineer/Scientist’s job to find the most efficient solution to the problem, while keeping all influencing factors in mind. I find it interesting and engaging to read about widely used algorithms and it helps me design my own code and benefits my thinking process.
Research Internship at Centre of Excellence on Safety Engineering & Analytics at Indian Institute of Technology, Kharagpur (IIT-KGP).
I have done a Research Internship, (6 months, part time), April – Sept, 2022 at Centre of Excellence on Safety Engineering & Analytics at Indian Institute of Technology, Kharagpur (IIT-KGP). During this time I got an opportunity to explore:
- Learned and experienced how to approach a research study and work towards attempting a solution to a practical real-life problem
- Explored topics such as Artificial Neural Networks, Regression Models, Classification Models and their application to real-life datasets
- Learned various classification methodologies including Multivariate Regression, Logistic Regression, KNN Imputation, Gradient Descent, and Artificial Neural Network
- Learned mathematical functions like cost and loss functions, activation functions like ReLU and sigmoid, et cetera
- Learned to use Python libraries such as Matplotlib, Pandas, Numpy, Sklearn, Seaborn, etc were used.
I worked on the dataset US Accidents (2016-2021), which has more than 2.8 million records and 49 attributes. The dataset was first analysed with visual representation to understand the different attributes better. Then classification was used to predict severe accident-prone situations. This project I completed Project: Data Analysis of US Road Accidents (2022), brief is as below.
The dataset is taken from Kaggle. It is a countrywide traffic accident dataset, which covers 49 states of the United States of America. The accident data are collected from February 2016 to December 2021, using multiple APIs that provide streaming traffic incident or event data. These APIs broadcast traffic data captured by a variety of entities, such as the US and state departments of transportation, law enforcement agencies, traffic cameras, and traffic sensors within the road-networks. The dataset contains 2,845,342 (2.84 million) rows and 49 columns
The main objective of the work was to show various important relations between attributes of the events. Data points which have the most influence have been highlighted. The dataset was first analyzed with visual representation to understand the different attributes better. Binary classification was then done by implementing an artificial neural network to predict severe accident-prone situations. Python libraries used included Pandas, Scikit-learn, NumPy, Matplotlib.pyplot, Matplotlib.seaborn, and Matplotlib.pylab.
Fellowship at Codebozu, a Cornell University eLab venture, USA
I have done a Fellowship for 6 weeks, part time in 2022 with Codebozu, Cornell University, USA, an Edutech start-up. I gained insight into Python modules like OpenCV, NumPy, and matplotlib. Learnt how to bring out the different attributes of an image and how to use kernels/convolution matrix to build filters, shaders, blurs, etc. I completed a project where I built a unique Image Filter for Image Manipulation. I explored the modules of OpenCV, NumPy, and Matplotlib to facilitate image manipulation. I learnt how to bring out or enhance different attributes of an image. I especially found the concepts of kernels and filters very interesting. The project allowed me to work along specific guidelines to make my own image filter, thus increasing my learning speed greatly.
Self driven Project Web Alert
During my school in 2021, I completed a project ‘Web Alert’, an application to help students organise and coordinate their online classes, notices, and study material. It has a web crawler that goes through the access-controlled school website looking for new notices, zoom links, etc. The user interface is simple and hassle free to facilitate easy access to class links. The program has a notification/alert system, which reminds users of their upcoming classes and notifies them when there are new notices. Project was done using Python and vast array of languages and technologies were used to build a web crawler, Tkinter window, database for easy access, log in mechanism, notification system, and attendance record.
I am learning Artificial Intelligence and Machine Learning using Python from India’s leading learning institute Codingal during 2021-22. During this time I am exploring topics such as Data Processing and Pre-processing, Classification, Regression, Deep Learning, Artificial Neural Networks, Convolutional Neural Networks, etc. Completed various projects like Handwritten Digit Recognition, House Price Prediction, Petrol Consumption Prediction, Malignant-Benign dataset classification, churn modelling, etc.
Project: Digit Identification (2022)
It is an application of multi class classification. First, the appropriate Python libraries are imported and the dataset is generated. Then the data is visualized with the help of graphs. The data is then split into multi class parts and logistic regression is applied. Lastly, a confusion matrix is displayed to better understand the accuracy of the model.
Project: Movie Recommendation System (2022)
A movie recommendation system was built using multi-class classification, where a movies/ratings dataset was taken, movies were appropriately assigned scores on different genres (0 or 1). Input was taken from the user in the form of movie names and their corresponding ratings. A metric – recommendation score – was taken to sort the suitable recommended movies. The top few movies were printed as output. Python libraries used were Pandas, NumPy, and Matplotlib.
Project: Customer Churn Prediction (2022)
Customer dataset was taken and analyzed, and unrequired attributes were dropped. Important features were visually represented to understand their effects better. The attributes were then altered to have numeric scores between 0 and 1. The resultant dataset was broken into training and testing, and the training dataset was fit into a neural network with 2 hidden layers, both having activation function ReLU, and the output layer having activation function Sigmoid. Finally, a confusion matrix was taken to properly visualize the accuracy of the model.
In 2019 I have completed a course on Arduino from MakersLoft where I learnt the usage of various sensors like temperature, light, ultrasonic, gyro, sound, humidity, etc. used motors, LCDs, speakers, and LEDs. Also learnt C++ for Arduino IDE
CodeBozu certificateCodeBozu certificate