Babandeep Singh

Chicago, IL – ***** +1-312-***-**** adgv8s@r.postjobfree.com LinkedIn GitHub EDUCATION

MS in Business Analytics, University of Illinois at Chicago GPA: 4/4 Expected: 05/2021 Courses: Statistical Models & Machine Learning Methods, Advanced Database Management, Data Mining for Business, Advanced Textual Analysis, Analytics for Big Data, Advanced Predictive Models, Social Media and Network Analysis MS in Mechanical Engineering, University of Illinois at Chicago GPA: 3.44/4 08/2015 – 12/2016 Courses: Design and Analysis of Experiments, Advanced Mathematical Methods Bachelor’s in Mechanical Engineering, University of Mumbai 08/2010 - 06/2014 SKILLS

Language and Software: Python C++ SQL Jupyter Notebook R Tableau Looker MATLAB CAD AWS Data Science: Pandas NumPy SciPy Data Visualization (Seaborn, Matplotlib, Plotly) TensorFlow Keras PyTorch Scikit-Learn Machine Learning: Natural Language Processing (nltk, SpaCy) Neural Networks (CNN, RNN) unsupervised learning (clustering, PCA)

Regression (Linear, logistics, lasso, Ridge) Classifiers (SVM, K-NN, Decision trees) Generative (Naïve Bayes, LDA) PROFESSIONAL EXPERIENCE

Research Assistant, University of Illinois at Chicago 05/2020 – Present

• Applying machine learning and NLP techniques (NER, SpaCy) for information retrieval (question-answering) modeling.

• Designed a metric/model that can evaluate partial correctness, as missing just one word from the right annotation could make the classification incorrect.

• Implemented Out-of-core classification, State-of-the-art Flair and CRFsuite models, and Neural networks to validate the correctness based on word embeddings, contextual embeddings via LSTM/GRU and familiarity coefficients. Product Engineer, Riddell Sports 04/2018-09/2019

• Developed products such as next generation State-of-the-art 3D-printed NFL helmets, ran FEA simulations and stress-tests on all 3D models before prototyping.

• Performed market analysis for feasibility, cost, and product placement against competing products and created functional A/B testing to ascertain remedies.

Mechanical Engineer, Freedom Motors USA Inc 03/2017-04/2018

• Perform initial and ongoing design reviews, identified bottlenecks, and developed contingency plans and built forecasting of raw material requirements to bring those deficiencies down.

• Developed CAD drawings and reviewed drawings from junior team members for accuracy. PUBLICATIONS

Chat Generator

Leveraging Recurrent Neural Networks to learn syntactic and semantics of WhatsApp personal chat to try to reduce screen time and a step towards easing Nomophobia and Phantom Vibration Syndrome.

- Published in The Innovation – [click here]

Sentiment Analysis – Amazon Foods

Created a model to understand what products within amazon foods are performing better amongst the words of the customers, and how positive and negative to reviews are using VADERSentiment library and WordClouds.

- Published in The startup – [click here]

Effect of Outliers in classification and Parameter Optimization Optimized Statistical models such as k-nearest neighbors (k-NN) and Decision trees to show how outliers in Hepatitis C Virus (HCV) in Egyptian patients can skew the judgment to standardizing assessment of histology among pathologists. Also, discussed how the outliers can be treated.

- Published in Towards Data Science – [click here] [click here] Resume summarization and matching

A vanilla approach to summarize job description and extract keywords to tailor resume and report the document matching to establish a confidence in job application. Reduce reading time and increasing in depth understanding of the job responsibilities.

- Published in Towards Data Science – [click here] ACADEMIC PROJECTS

Image Captioning 08/2020 – Present

Working on learning to generate Captions for an unseen image by training Convolutional Neural Networks and Recurrent Neural Networks. The model is expected to generate caption mimicking general population text to describe an image or experience. Kaggle Competition- Fifa2019 Dataset 01/2020 - 05/2020 Developed a classifier for investors to wish to own a soccer team using machine learning which is better than random guessing. Utilized Classifiers and feature engineering to get accuracies up to 94%. Market Segmentation 01/2020 - 04/2020

Clustered market segments for targeted campaign runs using unsupervised machine learning methods (k-means, hierarchical, DBSCAN

& k-medoids) & skillfully leveraging classification trees to find the target customers for right campaign in R. Credit Default Detection 01/2020 - 04/2020

Created ensemble (Random Forest) and boosting (XgBoost, Gradient Boosting) models to predict loan defaulters on a lending club data with 87% of accuracy. Utilized Sampling methods to mitigate effects of imbalanced classes. Yelp Sentiment Analysis 01/2020 - 05/2020

Created models for sentiment analysis using AFINN, Bing & NRC dictionaries for yelp restaurant reviews. used Naive Bayesian classifier, regression (decision trees, logistic) & classifiers (SVM, k-NN, Random Forest,) with accuracy of 87%. CERTIFICATIONS

• AWS Machine Learning – udacity 2020

• Using Databases with Python – Coursera 2019

• Python Data Structures- Coursera 2019

• Competitive and Market Analysis for Product Managers – LinkedIn 2019

• Product Management and Development Foundations – LinkedIn 2019

• Six Sigma: Green Belt – LinkedIn 2019

Contact this candidate