Scenario

You are working as a data scientist for a renowned mood$-$changing center named Bright$-$Hope. You have been charged with identifying patients who have mood disorders. Based on available data from previous patients, your manager has asked you to prepare a thorough report that outlines your findings along with a $3-$minute presentation for the CEO. Good luck!

Problem Statement

Medical professionals work hard to improve patients$$ lives. The advancement of statistical analysis and machine learning and applying these concepts in the medical field, has allowed medical staff to leverage the expertise of data scientists to predict the onsets of diseases and medical emergencies, and classify diagnosis, to just name a few.

You will build several prediction models to identify patients who suffer from mood disorders, using the expertise gained in this course: importing libraries, loading data into data frames, manipulating and wrangling data, exploring and visualizing data, writing functions, feature engineering, developing models, and finally evaluating and comparing models. The goal is to improve patients' lives.

Data

You will be using the Mood Disorder Dataset Download Mood Disorder Datasetavailable through canvas. The dataset consists of many features and the target class $($Diagnosis$).$ As part of your exploratory data analysis, you need do is explain the features and list the unique values in each feature.

Part $1$ Prediction models Report

This part is a group effort. The groups are assigned randomly unless you have emailed the professor about your team preference.

In this part, you will create a Jupyter Notebook using the dataset described above, which is available from Canvas. You will be doing these steps:

Import Libraries

Load dataset

Exploratory Data Analysis

Remove identifiable features to preserve privacy.

Data Dimension

Data Types

Summary Statistics

Correlation plots

Data Distribution $($plot features against Target variable$)$

Data Pre$-$Processing and Wrangling

Check for Missing Values

Missing values imputation $($if needed$)$

Check for Duplicate Data

Feature Engineering

Outliers

Categorical Data Encoding

Feature Scaling

Build functions.

Models Building:

K$-$Nearest Neighbors

Decision Tree

Random Forest

Extra Tree Classifier

Models Evaluation and Comparisons

K$-$Fold Cross Validation

Confusion Matrix

Accuracy

Precision

Recall or Sensitivity

Specificity

F$1-$score

Part $1$ Deliverable is a PDF version of your Notebook submitted once to the group project drop box. In Jupyter Lab, File $->$ Save and Export Notebook AS $->$ PDF$.$ If you get an error saving as PDF$,$ save it as HTML$,$ then open HTML file and right click and print to PDF$.$