/Personal_Kernels

Data science projects worked on by Matt Burnham

Primary LanguageJupyter NotebookMIT LicenseMIT

1. VGG16_ImageNet_4C_Blood_Classification_ver4.ipynb

My Capstone Project: LEUKOCYTE CELL CLASSIFICATION FOR RAPID SCREENING OF BLOOD CANCER
Completed 13th March 2020

Accurate classification of morphological classes of leukocyte cells that are prevalent in the detection of acute myeloid leukemia (AML),being greater than the accuracy and efficiency of a human.

Q. Whether computer vision/ deep learning / a convolutional neural network has better accuracy and is more time efficient than a human being in the task of classifying blood cells that are prevalent in blood cancers such as AML.

2. Mini-Project3_customer-segmentation_v3.ipynb

Dataset: Kaggle – “sales_data_sample”: A Toy wholesaling business

My Mini Project: Customer Segementation of a toy wholesaling business
Completed: Feb 2020
Feature engineering for K-Means clustering using RFM - Recency , Frequency, Monetary Value