Scribd
Upload
3 views1 page

03 02 DatasetAnalysisProblem

The document outlines a project for a Master's program focused on analyzing a breast cancer dataset. It includes goals such as dimensionality reduction, clustering, and classification using techniques like K-means, PCA, SVM, and RF. A report summarizing the methodology and findings is required by May 4, 2026.

Uploaded by

arnau.lacambra.troya
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
3 views1 page

03 02 DatasetAnalysisProblem

The document outlines a project for a Master's program focused on analyzing a breast cancer dataset. It includes goals such as dimensionality reduction, clustering, and classification using techniques like K-means, PCA, SVM, and RF. A report summarizing the methodology and findings is required by May 4, 2026.

Uploaded by

arnau.lacambra.troya
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

Máster en Tecnologías Mecánicas

Ciencia de Datos en Ingeniería Mecánica

Problem 3.2: Analysis of a Large Dataset


Dimensionality Reduction, Clustering, and Classification

Goals:

• Describe and analyze a large dataset of breast cancer samples


• Assess the relative importance of data features
• Reduce the dataset dimensionality
• Group samples using unsupervised techniques
• Train and evaluate a supervised classification algorithm to identify benign and
malignant samples.
Data source: Breast Cancer Wisconsin (Diagnostic) dataset
Tasks:
1. Describe the dataset, including features and targets.
2. Using only the features (unsupervised classification), demonstrate the application
of K-means clustering. Assess the most likely number of classes in the dataset from
statistical considerations.
3. Apply PCA to the features to reduce the space dimensionality. Make the selection of
how many PCs to use and justify your decision. Show the scree plot and the score
plots resulting from your selection of PCs.
4. Repeat the application of K-means using your selected PCs instead of the original
features. Discuss the observed changes.
5. Using both the features and the targets (supervised classification), demonstrate the
application of SVM and RF to predict the ‘benign’ or ‘malignant’ status of an unseen
sample. Make sure you compute model performance metrics.
Summarize your methodology and key findings in a report of maximum 5 pages.
Deadline: 4 May 2026

You might also like