SHABNAM HASAN

Masters of Applied Computer Science (Sept 2018 - June 2020)
Concordia University, Montreal, Canada

ABOUT ME

I am a Software Engineer, who takes pride in building models that translate data into insight. My immense interest in the field helped me in acquiring in-depth knowledge through theory courses and project development.
In my free time, I like to spend time with my family and friends, do painting, watch movies(sci-fi and horror), listen to music, visit places, workout and what not.

Recent activity

WON Co-operathon-2020 Dialogue challenge(Nov 2020) click here
Our application Cratos offers 1-1 remote physiotherapy along with services that use computer vision to help patients. Our MVP is in progress.

Attended DLRL summer school(3-7 August 2020)click here

Home Fitness Application(AILaunchLab)(May - August 2020)
Build a “Home Fitness Application” model from scratch to identify the workout type performed by the user and correct them in real-time.
Work involves:
Brainstorming on finding, cleaning, preprocessing the right dataset,
Achieve pose recognition and correction through LSTM model and geometric heuristic,
Maintain a high performance in a virtual environment.

Working on Autoencoders click here
Experimenting with autoencoders to achieve dimensionality reduction.

Skill set

Python, Django, Java, kotlin,
ML algorithms, Deep NN, CNN, RNN, LSTM, Image Analysis, computer vision, autoencoders, NLP, Pytorch, tensorflow ,scikit-learn, matplotlib, nltk, numpy, pandas, spark, AWS for ML,
RestAPI, postman, Git, Bitbucket, Jira, Agile methodology

Projects

Recommendation System using Collaborative Filtering click here
Perform collaborative filtering using Memory and model based to train a Recommendation system on MovieLens data. Uses item based and user based along with SVD, knn and ALS technique to perform the task.

Machine Learning
The course project involves collecting and preparing datasets(from source UCI repository, kaggle challenge and the UoT CIFAR dataset website) for training supervised learning algorithms(Classification and Regression models) to predict the output of future unseen inputs.
Pipeline used:

Performing hyperparameter search, evaluating (using F1 score, precision, recall, RMSE, r2 score) and summarizing performance by generating plots.

Image Processing click here
Reproduce the research paper written by Reza Zahiri, “Motion Estimation in Ultrasound Images Using TDPE”, using MATLAB to measure displacement of tissue in elastography images under compression to infer its elastic property.

Artificial Intelligence click here
Teaming with a classmate to build an efficient adversarial search algorithm to play the game Double Card. We implemented minimax and alpha beta pruning along with our heuristics, to optimize the search. Our AI beat 85% of the students during the class tournament.
The 2nd project includes building a Spam\Ham classifier using Naive-Bayes classifier to sort emails into SPAM and HAM classes. We started by cleaning the document by removing stop words, removing most and least frequent words, stemming and then creating the inverted index and performing the classification. Confusion matrix was used to see the result.

Information Retrieval and Web Search using Web Crawling & aFinn sentiment analysis
Build a web search engine, which displays the top 10 most relevant URLs to the user ,based on the cosine similarity and the aFinn score(sentiment analysis).
Implemented Boolean retrieval and Ranked retrieval model, on the inverted index, formed using SPIMI algorithm ,using Python and its library NLTK and BeautifulSoup.

Comparative study of programming language
3 Individual projects were given in 3 different languages, namely python, clojure and erlang in a very short duration. With strong learning skills, I learnt clojure and erlang from scratch and built the project efficiently in the time frame.

Work Experience

Software Engineer at Wipro Technologies (sept 2015 - july 2018), Bangalore, India.

Worked as Android app developer for an inventory management application for ESAB IoT.

In recognition of performance and continued contribution,promoted to higher band and awarded 5-star rating for consecutive quarters.

Worked as Automation Test Engineer for Charles Schwab Bank and Walmart Retail link.

Tools Used: Android Studio, Selenium, HP QTP, Eclipse, See Test Automation, SOAPUI

Languages Used: Java, VB script, Groovy, Python

Database Used: MongoDB, MySQL, DB2