Category: healthtech

  • The Power of AIHealth: Comparison of 12 ML Breast Cancer Classification Models

    The Power of AIHealth: Comparison of 12 ML Breast Cancer Classification Models

    Contents: BC Dataset Conventionally, the Breast Cancer Wisconsin (Diagnostic) Data Set has been used to predict whether the breast cancer is benign or malignant. Features were computed from a digitized image of a fine needle aspirate (FNA) of a breast mass. They describe characteristics of the cell nuclei present in the image. The dataset can […]

  • A Roadmap from Data Science to BI via ML

    A Roadmap from Data Science to BI via ML

    This post describes a Data Science (DS) roadmap, with relevant business applications. It has been written for aspiring data scientists, technical experts who work with data scientists, data-driven technology stakeholders, or anyone interested in learning about what DS is and what it’s used for. Why DS: The average base salary of a data scientist in […]

  • A Comparison of Scikit Learn Algorithms for Breast Cancer Classification – 2. Cross Validation vs Performance

    A Comparison of Scikit Learn Algorithms for Breast Cancer Classification – 2. Cross Validation vs Performance

    Featured Photo by Tara Winstead @ Pexels. This post is a continuation of the previous breast cancer (BC) study focused on a comparison of available Scikit-Learn binary classifiers (Logistic Regression, GaussianNB, SVC, KNN, Random Forest, Extra Trees, and Gradient Boosting) in terms of cross validation and model performance/scalability scores. Contents: Let’s set the working directory YOURPATH […]

  • A Comparison of Binary Classifiers for Enhanced ML/AI Breast Cancer Diagnostics – 1. Scikit-Plot

    A Comparison of Binary Classifiers for Enhanced ML/AI Breast Cancer Diagnostics – 1. Scikit-Plot

    The goal of this post is a comparison of available binary classifiers in Scikit-Learn on the breast cancer (BC) dataset. The BC dataset comes with the Scikit-Learn package itself. Contents: Data Analysis Let’s set the working directory YOURPATH import osos.chdir(‘YOURPATH’) os. getcwd() and load the BC dataset from sklearn import datasetsdata = datasets.load_breast_cancer() with the […]

  • ML-Assisted ECG/EKG Anomaly Detection using LSTM Autoencoder

    ML-Assisted ECG/EKG Anomaly Detection using LSTM Autoencoder

    Featured Photo by Luan Rezende Automatic detection and alarm of abnormal electrocardiogram (ECG aka EKG) events play an important role in an ECG monitor system; however, popular classification models based on standard supervised ML fail to detect abnormal ECG accurately. In this project, we implement an ECG anomaly detection framework based on the recently proposed […]

  • ML/AI Breast Cancer Diagnosis with 98% Confidence

    ML/AI Breast Cancer Diagnosis with 98% Confidence

    We demonstrate the importance of hyperparameter optimization (HPO) for enhancing ML prediction accuracy. Specifically, we will focus on the Random Forest Classifier (RFC) as an ensemble of decision trees. RFC is a supervised ML algorithm that has been applied successfully to the BC binary classification. 

  • Cloud-Native Tech Autumn 2022 Fair

    Cloud-Native Tech Autumn 2022 Fair

    Let’s dive deeper into the cloud-native tech trends and features to follow in Q4 2022 and beyond. Contents: Markets Services Serverless Cybersecurity DevSecOps ML/AI/IoT Use-Cases Events Training Explore More Infographic

  • Breast Cancer ML Classification – Logistic Regression vs Gradient Boosting with Hyperparameter Optimization (HPO)

    Breast Cancer ML Classification – Logistic Regression vs Gradient Boosting with Hyperparameter Optimization (HPO)

    Breast Cancer (BC) is the leading cause of death among women worldwide. The present study optimizes the use of supervised Machine Learning (ML) algorithms for detecting, analyzing, and classifying BC. We compare Logistic Regression (LR) against Gradient Boosting (GB) Classifier within the Hyperparameter Optimization (HPO) loop given by GridSearchCV. We use the publicly available BC dataset from the University of Wisconsin Hospitals, Madison, Wisconsin, USA. Feature engineering yields 9 dominant features Results: AUC(LR)=0.987 > AUC(GBM)=0.975 It is clear that LR has the highest f1-score and POS.

  • A Comparative Analysis of Breast Cancer ML/AI Binary Classifications

    A Comparative Analysis of Breast Cancer ML/AI Binary Classifications

    This study is dedicated to #BreastCancerAwarenessMonth2022 #breastcancer #BreastCancerDay @Breastcancerorg @BCAction @BCRFcure @NBCF @LivingBeyondBC @breastcancer @TheBreastCancer @thepinkribbon @BreastCancerNow. One of the most common cancer types is breast cancer (BC), and early diagnosis is the most important thing in its treatment. Recent studies have shown that BC can be accurately predicted and diagnosed using machine learning (ML) technology. Our objective is to compare different supervised ML, deep learning (DL) and data mining techniques for the early detection of BC. The idea is to analyze BC data based on its characteristics and identify the effectiveness of clustering and classification instructions for analyzing and fitting various ML models. We tested the performance of ML models by looking at their accuracies, sensitivities, specificities, and other metrics. Results obtained with the best ML model with most dominant features included showed the highest classification accuracy (~99%), and the proposed approach revealed the enhancement in accuracy performances. These results indicated the potential to open new opportunities in the BC research.

  • ML/AI for Diabetes-2 Risk Management, Lifestyle/Daily-Life Support

    ML/AI for Diabetes-2 Risk Management, Lifestyle/Daily-Life Support

    In the last decade, the impact of Type-2 Diabetes (T2D) has increased to a great extent especially in developing countries. T2D is a common condition that causes the level of sugar (glucose) in the blood to become too high. T2D is responsible for very considerable morbidity, mortality. The objective of this project is to summarize recent […]