Mahesh Raja

Mahesh Raja

San Francisco Bay Area
2K followers 500+ connections

Join to view profile

About

Data science and analytics leader with 14+ years of experience driving product-led…

Experience

  • DoorDash Graphic

    DoorDash

    San Francisco Bay Area

  • -

    San Francisco Bay Area

  • -

    San Francisco Bay Area

  • -

    Austin, Texas Area

  • -

    Austin, Texas Area

  • -

    Austin, Texas Area

  • -

    Dallas/Fort Worth Area + New York

  • -

    Bryan/College Station, Texas Area

  • -

    Dallas/Fort Worth Area + New York

  • -

    Chennai Area, India

Education

Projects

  • StackGen - Python package for stacked generalization

    Stacked generalization or 'stacking' has become a staple technique in Kaggle competitions (and other data science competitions) among top competitors. This technique is used to combine multiple machine learning models using a meta-learner and can greatly improve generalization error. However, the implementation of stacked generalization is rather tedious and confusing for beginners and pros alike. The StackGen package aims to hide the complexity behind this technique and provides an easy…

    Stacked generalization or 'stacking' has become a staple technique in Kaggle competitions (and other data science competitions) among top competitors. This technique is used to combine multiple machine learning models using a meta-learner and can greatly improve generalization error. However, the implementation of stacked generalization is rather tedious and confusing for beginners and pros alike. The StackGen package aims to hide the complexity behind this technique and provides an easy, familiar api for users to perform stacking. The idea is based on the seminal paper 'Stacked Generalization' by David Wolpert (Ref: Wolpert, David H. "Stacked generalization." Neural networks 5.2 (1992): 241-259).

    https://github.com/mraja9/StackGen

    See project

  • Training a smartcab to drive using Q-Learning (reinforcement learning)

    In this project, a self-driving agent was trained using the Q-Learning algorithm to reach its destination in allotted time without accidents. The simulated world follows the US driving rules and has other cars in the environment along with traffic lights at intersections. The trained smartcab achieved a rating of A+ for safety (i.e. agent commits no traffic violations, and always chooses the correct action) and A for reliability (i.e. agent reaches the destination on time for at least 90% of…

    In this project, a self-driving agent was trained using the Q-Learning algorithm to reach its destination in allotted time without accidents. The simulated world follows the US driving rules and has other cars in the environment along with traffic lights at intersections. The trained smartcab achieved a rating of A+ for safety (i.e. agent commits no traffic violations, and always chooses the correct action) and A for reliability (i.e. agent reaches the destination on time for at least 90% of trips).

    https://github.com/mraja9/ReinforcementLearning_smartcab

    See project

  • Classification of CIFAR-10 images using TensorFlow

    Built a convolutional neural network using TensorFlow to classify images from the CIFAR-10 dataset. The network includes convolution and max pool layers, a flatten layer, fully connected layers and an output layer.

    https://github.com/mraja9/image_classification

    See project

  • Data Warehousing

    • Designed and developed a data warehouse for the retail store chain Dominick's Finer Foods using STAR schema data model.
    • Used the Microsoft Business Intelligence Stack – SQL, SSIS, SSAS, SSRS to perform Extract Transform Load processes, build OLAP cubes and develop reports to answer critical business questions about discounts' effectiveness, shelf space utility and sales trends.
    • Performed exploratory data analysis using Tableau and Excel to gain better insight into the data and…

    • Designed and developed a data warehouse for the retail store chain Dominick's Finer Foods using STAR schema data model.
    • Used the Microsoft Business Intelligence Stack – SQL, SSIS, SSAS, SSRS to perform Extract Transform Load processes, build OLAP cubes and develop reports to answer critical business questions about discounts' effectiveness, shelf space utility and sales trends.
    • Performed exploratory data analysis using Tableau and Excel to gain better insight into the data and detect outliers and anomalies.

  • Health Care Data Mining and Predictive Analytics

    Capstone project for the graduate Data Mining course at Texas A&M University -
    • Predicted the risk of mortality of patients by building prediction models using algorithms such as Multiple Linear Regression, Random forest, Naive Bayes, K-Means, and K-NN and evaluated models using classification matrices.
    • Conducted exploratory data analysis using Tableau for health care data.
    • Performed dimension reduction using Principal Component Analysis (PCA), treated missing data and outliers…

    Capstone project for the graduate Data Mining course at Texas A&M University -
    • Predicted the risk of mortality of patients by building prediction models using algorithms such as Multiple Linear Regression, Random forest, Naive Bayes, K-Means, and K-NN and evaluated models using classification matrices.
    • Conducted exploratory data analysis using Tableau for health care data.
    • Performed dimension reduction using Principal Component Analysis (PCA), treated missing data and outliers through imputation techniques.

View Mahesh’s full profile

  • See who you know in common
  • Get introduced
  • Contact Mahesh directly
Join to view full profile

Other similar profiles

Explore more posts

Explore top content on LinkedIn

Find curated posts and insights for relevant topics all in one place.

View top content

Others named Mahesh Raja in United States

6 others named Mahesh Raja in United States are on LinkedIn

See others named Mahesh Raja