Project 1: Object-Detection Using Tiny Yolo in R
In this project I created a quick object detection model with a simple code block leveraging Tiny Yolo. Yolo (You Only Look Once) object detector is often cited as being one of the fastest deep-learning based object detectors.
*Created a PoC concept to augment organizations sales pipeline *Later in collaboration with the datascience team created a realtime object detection module
Project 2: Logistic-Regression-Shiny-App
The aim of this project was to create an app that can help its users to perform quick logistic regression routines without having to write numerous lines of codes.This app allows its users to generate Logistic Regression and SVM based models using R’s famous CARET (Classification and Regression Training) Logreg and SVM packages.
Note: In order to use this application your data should be in binary format (0,1) in csv format
Project 3: Time Series Forecasting with Shiny
This shiny app was create in collaboration with Rjshanahan where users can upload their own CSV with single or multiple daily time series. The user interace allows users to compare fitted time series models and forecasts with several algorithms including:
*Line of best fit (regression) *Moving average *Exponential smoothing (simple & Holt-Winters) *ARIMA with Fourier Transform *TBATS *Hybrid forecast ensemble model
Project 4: Text Analytics Shiny App
This text analytics app allows even a rookie NLP practioner to parse information out of spreadsheets or text documents allowing them to answer a series of cross-cutting questions. In this project contains customized R’s “tidytext” and “plotly” packages and UVM’s “LabMTsimple” sentiment dictionary for textual and sentiment analysis and various intervention points for users to subset the data, ask it guided questions and produce exploratory data visualizations.
Project 5: HR Analytics Employee Attrition
Employee retention plays an important role in the success of any organization and the effectiveness of its HR department. To make this experiment even more interesting, I have used the Permutation Feature Importance widget that can be used to compute the variable importance given the scores based on the trained model and test data. This model was consumed via webservice as an Azure Webservice Plugin.
This model helped me to identify attrition indicators at the early stage and reduced the attrition by almost 20% resulting in cost reduction related to new employee aquisition and trainings.
Project 6: PyCaret-Gradio-Heart Stroke Prediction-App 🔮
This app is created using PyCaret low code machine learning library and Gradio an open source and easy-to-use UI for your ML model, function, or API with only a few lines of code. Integrate directly into your Python notebook, or share a link with anyone.
The video belows shows a Heart Failure prediction (1 = Stroke & 0 = No Stroke) based on the selections made by the users. Gradio and PyCaret can be used to create compelling Ml PoCs.
Data Visualization Using R
Data Chat Shows & Webinars
Badges and Accreditations
