Ashish Bansal Machine Learning Engineer at Vanguard

About Me Projects Blog Resume

Who am I

me

Innovative and motivated Data and Software engineer with two years of experience in "object-oriented design, Machine learning, Big Data technologies, deep knowledge of data structure and algorithms". My detailed expertise can be found in Resume

Besides coding, I always explore the recent trends in the fitness industry and being a fitness enthusiast, I want to contribute to this industry in the future.

I pick something new to learn every year and new adventures to complete. Life is officially too short for my long list of things to learn.

Code
Frameworks
Tools

Featured Projects

code
Predicting Credit Card Defaulters
  • Skills: Apache spark, pyspark, Python

    • Predicted credit card defaulters beforehand analyzing there payment history attributes such as average pay, average bill amount for the credit card companies to monitor there expenses and there payments through python script written using pyspark and pyspark.mllib.

    Source Code
    code
    iNEWS
  • Skills: AWS, Python

    • Mobile application that uses a News API to fetch the latest news from sources including CNN and the New York Times and deliver it to the user; whether or not user is connected to the Internet.

    Demo Source Code
    code
    ANALYZING NYPD COMPLAINT DATA (HISTORIC)
  • Skills: Hadoop, Apache Spark, SQL, Python, Pandas

    • Analyzed NYPD Complaint data to uncover hidden patterns, unknown correlations, crime trends and other anomalies. Generated hypothesis based on revelations by correlating with datasets like Weather, census and employment data.

    Source Code Project Report
    code
    YELP DATASET ANALYSIS
  • Skills: Python, Apache Spark, PigLatin, Hive, Map-Reduce, Docker

    • Performed some basic statistics like summarizing reviews by city and category, ratings of businesses around University of Wisconsin-Madison, based on number of reviews on the dataset by executing scripts written in PigLatin and ApacheSpark

    Source Code Project Report
    code
    CLASSIFIERS TO IDENTIFY TWITTER ACCOUNTS AS BOTS OR NOT BOTS
  • Skills: Python, Scikit-learn, Pandas, Numpy, scipy

    • A Built four classifiers (Multinomial Naive Bayes, Decision Trees, Logistic Regression and Random Forest) using python libraries pandas to train our model to predict twitter accounts as bots.Compared model’s accuracy from different classifiers we built.

    Source Code Project Report
    code
    ANALYZING TRAITS SHARED BETWEEN TWO TWITTER USERS
  • Skills: Python, Scikit-learn, watson_developer_cloud, Watson Personality Insights API

    • Built a fully functional, Watson-powered application using Python to interact with the Twitter API and IBM's Personality Insights API in order to analyze traits shared between two Twitter users and Displayed the top 5 personality traits shared between two Twitter users

    Source Code
    web counter