Hi there!
My name is Farisayi (fah-rii-sai-ee). I am a Natural Scientist turned Data Scientist that loves data and problem-solving. This website is a portfolio (my technical journal if you will) of projects I have completed, skills and technics I feel inclined to wax eloquent about.
Welcome! l hope you enjoy your visit…
Latest Projects
GeoAI Aquaculture Pond Identification using Satellite data (Sentinel-1 + Sentinel-2 bands)
A machine learning solution for the Zindi GeoAI Aquaculture Pond Identification Challenge, detecting aquaculture ponds from Sentinel-1 radar and Sentinel-2 optical satellite data. My solution ranked 253/1314, generalising cleanly from a clean training period to a blind, cloud-affected test period. Continue reading GeoAI Aquaculture Pond Identification using Satellite data (Sentinel-1 + Sentinel-2 bands)
PART 2 - Product Sales Trend Analysis for Black Friday week
In this project, I build a interactive R Shiny Dashboard to present Sales Analysis Trends of Thanks giving week of 2021, Insights and Business Recommendation Continue reading PART 2 - Product Sales Trend Analysis for Black Friday week
PART 1 - Product text Classification (TF-IDF + Multinomial Naive Bayes)
In this project, I built a text classifcation model to predict categories of > 700 000 products item descriptions, leveraging the power of Multinomial Naive Bayes to handle large datasets. My model had an accuracy of 88% Continue reading PART 1 - Product text Classification (TF-IDF + Multinomial Naive Bayes)
Using Random Forest to Predict Customer Churn
Junior Data Scientist Job simulation at Boston Consulting Group (BCG), using feature engineering and machine learning models to determine the influence of price sensitivity on custormer churn. Continue reading Using Random Forest to Predict Customer Churn
See Projects for more