Learn practical skills, build real-world projects, and advance your career
# Jovian Commit Essentials
# Please retain and execute this cell without modifying the contents for `jovian.commit` to work
!pip install jovian --upgrade -q
import jovian
jovian.utils.colab.set_colab_file_id('17iO0rBs-gOFSUPbr6nd3IyfcKR-j2yYa')

Cancer Mortality rate prediction for US counties using feedfoward neural networks

Using data aggregated from the American Community Survey (https://census.gov), https://hclinicaltrials.gov, and https://cancer.gov.

Link of aggregated data can be found here: https://data.world/nrippner/ols-regression-challenge

ABSTRACT

This project aims to predict cancer mortality rates of US counties, using feedfoward neural networks. We'll first start by downloading and cleaning the data. Then we'll perform some exploratory data analyses to look for trends. We'll then build a model with the dataset, which takes a bunch of inputs and returns predictions for target death rate. Finally, we'll examine how how our model performed against The data

type of data:

tabular data

Model used:

 Regression with feed-forward neural networks

Introduction

Cancer is the second leading cause of death in the united states. Various types of cancer have been associated with modifiable risk factors like smoking, lack of physical activity, obesity, etc. Because modifiable risk factors have a very large association with the risk of developing cancers, half of cancer may be preventable (Colditz et al, 1996).

However, individual risk factors might not be the only modifiable risk factors for cancer. Parameters like poor access to health care, low income etc, should also be considered as potential indirect risk factors. It's no wonder, areas with better cancer detection centers, tend to have a lower rate of cancer mortality, thanks to early detection.

The dataset used in this project mainly contains population features of US counties, which will be fed into our model, in order to predict how these features might affect cancer mortality rates of US counties.

Note

A copy of the dataset on my local drive will be used. However, I'll leave the download url to the dataset below. if you wish to explore on your own machine, simply uncomment the download_url function and replace the read_csv path with your path

# Uncomment and run the commands below if imports fail
# !conda install numpy pytorch torchvision cpuonly -c pytorch -y
# !pip install matplotlib --upgrade --quiet
!pip install jovian --upgrade --quiet