• Customer Segmentation
  • Identifying Cluster Segments in Customer data for targeted Marketing .
  • Identified Initial set of KPI's & worked with huge transactional data of customers spread across multiple regions of U.S || Identified 4 Clusters in data || Cluster profiling in excel || Presentation prepared for the marketing incharge

• Insurance Fraud Detection(quick: See the notebook)
  • A Classification Problem to detect if the insuarnce claim is fraudulent or genuine.See the project

  Training Pipeline : Client Data Validation--> Data Ingestion-->Exploratory Data Analysis-->Data Preprocessing & Feature Engineering-->Model Training & Hyperparameter tuning || Prediction Pipeline :Client Data Validation--> Data Ingestion-->Exploratory Data Analysis-->Data Preprocessing & Feature Engineering-->Prediction || Deployment : Flask (developlment) & Google cloud Platform (Production) || Application logging maintaned

• Big Mart Sales(quick: See the notebook)
  • A Regression Problem to predict the sales of an item at a particular Outlet.See the project

  Training Pipeline : Client Data Validation--> Data Ingestion-->Exploratory Data Analysis-->Data Preprocessing & Feature Engineering-->Model Training & Hyperparameter tuning || Prediction Pipeline :Client Data Validation--> Data Ingestion-->Exploratory Data Analysis-->Data Preprocessing & Feature Engineering-->Prediction || Deployment : Flask (developlment) & Google cloud Platform (Production) Application logging maintaned

• Back Order Prediction(quick: See the notebook)
  • A Classification Problem to detect if a product could come short & Order has to be cancelled.(inventory management)See the project

  Training Pipeline : Client Data Validation--> Data Ingestion-->Exploratory Data Analysis-->Data Preprocessing & Feature Engineering-->Model Training & Hyperparameter tuning || Prediction Pipeline :Client Data Validation--> Data Ingestion-->Exploratory Data Analysis-->Data Preprocessing & Feature Engineering-->Prediction || Deployment : Flask (developlment) & Google cloud Platform (Production) Application logging maintaned

• Demand Forecasting(quick: See the notebook)
  • A Regeression problem to forecast the numbers of bikes to be rented at any hour of a day .|| Solution to Kaggle Challenge : Bike Rental Demand Forecasting SystemSee the project
• Sentiment Analysis See the notebook
  • Part I : A Classification problem to understand sentiment hidden in text & classify if it is positive or negative. (on IMDB movie reviews)|| Data http://ai.stanford.edu/~amaas/data/sentiment
  • Part II : webscraping the IMDB reviews from Imdb.com ,see notebook to collect data at large scale (done) || testing the efficiency of model & retraining (in process)

• Digit Recognizer android app See the project
  • An app that can identify the numbers from the image or when drawn by hand ||Technology : MNIST dataset,Tensorflow, Dart,Flutter
• Image Classification Web app See the project
  • A Flask app that classifies between cats & dogs images when uploaded by a user .||Technology : Tensorflow,Flask app
• Dog Breed detector android app (in process)

You can reach me at ✨

• Write to me ([email protected])
• DM me
• Follow me

©️ for octocat Octodex