Search Results for: --------

During this project I analyzed a dataset related to car purchases using a decision tree model. The goal of my project is to predict whether a user will purchase a car based on financial and age features. My dataset contains 5 columns including UserID, Gender, Age, AnnualSalery, and Purchased. I was able to train my decision tree model to get 90% of validation accuracy and 90.14% training accuracy.

I completed a decision tree analysis for MIS 2502 with Konstantin Bauman. I selected a data set detailing information about car sales, including the age, gender and income of prospective buyers and the outcome of their decision. By building the decision tree, patterns in the data were identified, allowing the prediction of whether a person will buy a car or not based on identifying features. While creating the decision tree, I tested various combinations of minimus split and depth settings to provide the most effective results and reduce the effects of over and underfitting.

Schoen_2502_PRO_points

car_data

This project involved analyzing a loan approval dataset using a decision tree algorithm to predict the approval status of loan applicants based on their financial features such as income, credit score, and loan amount. This project was part of my MIS2502 coursework. My role was to clean and analyze data, build a decision tree model, and interpret the results to gain insights into the approval process. Through this project, I identified the optimal minimum split for the decision tree, achieving high training and validation accuracy rates. I also explored nodes with extreme probabilities to uncover patterns in approved and rejected applicants. Additionally, I applied the model to real-world scenarios by predicting outcomes for individual applicants. This experience enhanced my understanding of decision tree algorithms, feature importance, and the role of predictive analytics in financial decision-making. It also improved my technical skills in data processing and model evaluation, as well as my ability to communicate complex findings clearly and effectively. Overall, this project demonstrated my ability to apply analytical tools to solve real-world scenarios.

The goal of this project is to provide MIS students with additional hands-on experience in data analysis and reinforce the concepts and methods covered in class.

For this project, students should find a new suitable dataset on the internet (that was never used in the class before) and apply Decision Tree analysis to build the prediction of the outcome variable. The process should be very similar to the regular assignment on Decision Trees (i.e., start with the same Jupyter Notebook) but applied to a new dataset.

annotated-diabetes_dataset.csv

annotated-Propointsproject.docx (1)

To complete this assignment, I sourced a dataset detailing patients and their health attributes that may or may not be correlated with having diabetes. This dataset proved more difficult for me to find a fitting minimum split than past ones have, but I was able to settle on using 50 because it maximized visibility as well as accuracy. I created the four scenarios with differing attributes to try and get an idea of both sides of the spectrum as far as likelihood of having diabetes. These scenarios were based on glucose levels, age, BMI, Diabetes pedigree function, and blood pressure. After analysis of the data, it became clear that certain attributes, like high blood pressure, BMI, and glucose levels, could help indicate a patient’s chance of having diabetes.

I imported a CSV file containing data from a bank and used Python in Jupiter Notebook to train a model that predicts customer outcomes. I produced a decision tree that evaluated the likelihood that a customer would leave the bank based on numerous demographics and metrics.

This project focuses on using a decision tree algorithm to analyze the salaries of NBA players based on their performance data. The main goal is to predict player salaries using key features such as their age, average minutes played, and win shares. By analyzing these features, we can understand how player performance impacts their earnings.

The project involves building and optimizing a decision tree model. First, we tested different values for the minimum split parameter to find the best balance between overfitting and underfitting. Then, we identified the highest and lowest salary nodes in the tree and analyzed the features that influenced them. Finally, we used the model to predict salaries for selected player data points, demonstrating how decision trees can provide meaningful insights into salary structures.

This project involved analyzing the 2025 IPL Auction dataset using Decision Tree modeling to predict whether a player is “Capped” or “Uncapped” based on their base price and winning bid. The analysis was conducted as part of an academic assignment for MIS2502, focused on applying data analysis techniques to a new dataset.

My role included data preprocessing, building the decision tree model, and extracting meaningful insights. I consolidated multiple team-specific datasets into a unified format, selected relevant features, and implemented the model using Python and Jupyter Notebook. Key deliverables included a detailed Word document summarizing the findings, a visual representation of the decision tree, predictions for test examples, and the Jupyter Notebook used for the analysis.

Through this activity, I developed a deeper understanding of Decision Tree algorithms, feature selection, and the trade-offs involved in tuning model parameters like minimum split. This hands-on experience reinforced my ability to handle real-world datasets, draw insights from data, and effectively communicate results through structured documentation and visualizations.

I analyzed a data set containing the sex, class, age, fare price, etc. of passengers on the titanic to predict whether or not they survived. I used a decision tree in Jupyter Notebook to classify the data and deciphered the best minimum split. Then I drew conclusions about the survivability of different passengers based on the factors detailed in the tree. Through this project I was able to further my understanding on decision tree analysis and apply it to a topic of my own interest.