Helping Spain - A Regression Problem
Helping Spain - A Regression Problem
The Load Shortfall Challenge
Embedded Files
This challenge was a regression problem that sought to help the Spanish government prevent load shortfall issues using data over a range of stated weather condition/features.
This challenge was a regression problem that sought to help the Spanish government prevent load shortfall issues using data over a range of stated weather condition/features.
The business application of this predictive capability would be leveraged in the energy market, using foreknowledge to predetermine energy purchase cost for optimum profitabilty.
The business application of this predictive capability would be leveraged in the energy market, using foreknowledge to predetermine energy purchase cost for optimum profitabilty.
The line chart image above shows the results of a base Linear Regression model and the Adaboost Regressor model, superimposed on the test target. It shows good prediction by the Adaboost model with no overfitting.
The line chart image above shows the results of a base Linear Regression model and the Adaboost Regressor model, superimposed on the test target. It shows good prediction by the Adaboost model with no overfitting.
On this project, I led a team of four scientists, working also on the deployment of the flask API.
Sentiments Analysis - A Classification Problem
Classification of Climate Change Tweets
An analysis of sentiments on subject that pertains to climate change helps businesses such as in the Tourism sector learn non-controversial ways to reach their market. The Energy Sector and Agriculture would also benefit immensely from this analysis.
An analysis of sentiments on subject that pertains to climate change helps businesses such as in the Tourism sector learn non-controversial ways to reach their market. The Energy Sector and Agriculture would also benefit immensely from this analysis.
The scatter diagram above shows the compares the metrics of train time and F1 score for all the tested models.
On this project, I worked with a team of four Data Scientists, and I was assigned to work on the deployment of the Streamlit API.
The model ranked 1st on a board of 15 teams.
Page updated
Google Sites
Report abuse