How I won my 42nd bronze medal on a Kaggle playground competition concerning smoking

Kaggle is a free data science company that hosts a number of machine learning competitions. I really get a lot out of entering Kaggle competitions because they provide me with a way to learn new skills in machine learning. I also like to give code reviews of my work because it enables me to review my work one last time before I go live. It just so happens that I have carried out a fair bit of work in an attempt to increase my score in Kaggle’s playground competition, season 3, episode 24. As it happens, I have carried out so much work in an attempt to improve my score that I have been awarded three bronze medals so far for the Jupyter Notebooks that I have created so far. I have therefore decided to discuss what I have done so far to improve my score on the Kaggle competition where I was required to predict the probability of a person being a smoker compared to several biological readings.

I endeavoured to improve the accuracy of my predictions by using a linear regression model that is composed of weights and biases. Weights are often used to account for uncertainty in the data values, for example if a given data point’s value is very uncertain then the weight for that data point is small. Bias refers to the error that is introduced by approximating a real-life problem, which may be complicated, by a…