Machine Learning In Healthcare
Expert-defined terms from the Professional Certificate in AI, Pediatric Care and Clinical Decision Making (Trinidad and Tobago) course at Stanmore School of Business. Free to read, free to share, paired with a professional course.
A/B Testing refers to a method of comparing two versions of a product, service,… #
Related terms include randomized controlled trials and experimental design. In the context of Machine Learning In Healthcare, A/B testing can be applied to evaluate the effectiveness of different prediction models or treatment strategies.
Accuracy is a measure of how close a machine learning model's predictions… #
Related terms include precision, recall, and F1 score. In healthcare, accuracy is crucial for diagnostic models and treatment recommendations.
Active Learning is a machine learning approach that involves actively sel… #
Related terms include semi-supervised learning and transfer learning. In the context of Machine Learning In Healthcare, active learning can be applied to image classification tasks, such as tumor detection in medical images.
Anomaly Detection refers to the process of identifying outliers or abn… #
Related terms include outlier detection and novelty detection. In healthcare, anomaly detection can be applied to electronic health records (EHRs) to identify inconsistent data or suspicious activity.
Artificial Intelligence (AI) refers to the development of computer systems</b… #
Related terms include machine learning, deep learning, and natural language processing. In healthcare, AI can be applied to medical imaging, clinical decision support, and personalized medicine.
Backpropagation is an algorithm used to train artificial neural networ… #
Related terms include stochastic gradient descent and optimization. In the context of Machine Learning In Healthcare, backpropagation can be used to train deep learning models for image classification tasks, such as tumor detection in medical images.
Bayesian Networks are a type of probabilistic graphical model that repres… #
Related terms include decision trees and randomrain forests. In healthcare, Bayesian networks can be applied to clinical decision support systems to predict patient outcomes and recommend treatments.
Bias #
Variance Tradeoff refers to the balance between bias (error due to simplifying assumptions) and variance (error due to noise in the data) in machine learning models, often used to evaluate the performance of regression models in healthcare. Related terms include overfitting and underfitting. In healthcare, the bias-variance tradeoff is crucial for model selection and hyperparameter tuning.
Big Data refers to the large amounts of unstructured and structured da… #
Related terms include data mining and data analytics. In healthcare, big data can be applied to predictive analytics, population health management, and personalized medicine.
Bootstrapping is a statistical technique used to estimate the distribu… #
Related terms include cross-validation and permutation testing. In healthcare, bootstrapping can be applied to confidence interval estimation and hypothesis testing.
Causal Inference refers to the process of drawing causal conclusions from… #
Related terms include confounding variables and instrumental variables. In healthcare, causal inference can be applied to clinical trials and outcomes research.
Classification is a type of machine learning task that involves predictin… #
Related terms include regression and clustering. In healthcare, classification can be applied to image classification tasks, such as tumor detection in medical images.
Clinical Decision Support Systems (CDSSs) are computer systems that provi… #
Related terms include expert systems and knowledge-based systems. In healthcare, CDSSs can be applied to diagnosis, treatment planning, and patient monitoring.
Clustering is a type of unsupervised learning task that involves grouping… #
Related terms include dimensionality reduction and anomaly detection. In healthcare, clustering can be applied to genomic data to identify genetic variants associated with diseases.
Convolutional Neural Networks (CNNs) are a type of deep learning model th… #
Related terms include recurrent neural networks and autoencoders. In healthcare, CNNs can be applied to image classification tasks, such as tumor detection in medical images.
Cost #
Sensitive Learning is a type of machine learning that takes into account the costs associated with misclassification or false positives, often used in healthcare to evaluate the effectiveness of diagnostic tests or treatment strategies. Related terms include cost-benefit analysis and cost-effectiveness analysis. In healthcare, cost-sensitive learning can be applied to resource allocation and priority setting.
Data Mining is the process of discovering patterns and relationships</… #
Related terms include data analytics and business intelligence. In healthcare, data mining can be applied to predictive analytics, population health management, and personalized medicine.
Deep Learning is a type of machine learning that uses artificial neura… #
Related terms include convolutional neural networks and recurrent neural networks. In healthcare, deep learning can be applied to image classification tasks, such as tumor detection in medical images.
Dimensionality Reduction is a technique used to reduce the number of f… #
Related terms include principal component analysis and t-distributed Stochastic Neighbor Embedding. In healthcare, dimensionality reduction can be applied to genomic data to identify genetic variants associated with diseases.
Electronic Health Records (EHRs) are digital versions of a patient's m… #
Related terms include personal health records and health information exchange. In healthcare, EHRs can be applied to clinical decision support, population health management, and personalized medicine.
Ensemble Methods are a type of machine learning that combines the predict… #
Related terms include bagging and boosting. In healthcare, ensemble methods can be applied to clinical decision support systems to predict patient outcomes and recommend treatments.
Evidence #
Based Medicine (EBM) is an approach to medical practice that emphasizes the use of best available evidence to guide clinical decision-making, often used in healthcare to evaluate the effectiveness of treatments and interventions. Related terms include clinical trials and systematic reviews. In healthcare, EBM can be applied to guideline development and quality improvement.
Feature Engineering is the process of selecting and transforming varia… #
Related terms include feature selection and feature extraction. In healthcare, feature engineering can be applied to genomic data to identify genetic variants associated with diseases.
Generalizability refers to the ability of a machine learning model to … #
Related terms include external validity and internal validity. In healthcare, generalizability is crucial for model deployment and clinical decision-making.
Health Information Exchange (HIE) refers to the electronic sharing of … #
Related terms include electronic health records and personal health records. In healthcare, HIE can be applied to clinical decision support, population health management, and personalized medicine.
Hyperparameter Tuning is the process of adjusting hyperparameters to o… #
Related terms include grid search and random search. In healthcare, hyperparameter tuning can be applied to deep learning models for image classification tasks, such as tumor detection in medical images.
Image Classification is a type of machine learning task that involves pre… #
Related terms include object detection and segmentation. In healthcare, image classification can be applied to medical imaging modalities, such as CT scans and MRI scans.
K-Means Clustering is a type of unsupervised learning algorithm that grou… #
Related terms include hierarchical clustering and density-based clustering. In healthcare, k-means clustering can be applied to genomic data to identify genetic variants associated with diseases.
Machine Learning is a type of artificial intelligence that involves train… #
Related terms include deep learning and natural language processing. In healthcare, machine learning can be applied to medical imaging, clinical decision support, and personalized medicine.
Medical Imaging is the process of creating images of the body</… #
Related terms include image analysis and computer vision. In healthcare, medical imaging can be applied to image classification tasks, such as tumor detection in medical images.
Natural Language Processing (NLP) is a type of artificial intelligence th… #
Related terms include text mining and sentiment analysis. In healthcare, NLP can be applied to clinical decision support, population health management, and personalized medicine.
Neural Networks are a type of machine learning model that uses artific… #
Related terms include deep learning and convolutional neural networks. In healthcare, neural networks can be applied to image classification tasks, such as tumor detection in medical images.
Overfitting occurs when a machine learning model is too complex an… #
Related terms include underfitting and regularization. In healthcare, overfitting can be prevented using techniques such as dropout and early stopping.
Personalized Medicine is an approach to medical practice that involves <b… #
Related terms include precision medicine and targeted therapy. In healthcare, personalized medicine can be applied to cancer treatment and genetic disorders.
Predictive Analytics is the process of using statistical models and ma… #
Related terms include descriptive analytics and prescriptive analytics. In healthcare, predictive analytics can be applied to population health management and clinical decision support.
Precision Medicine is an approach to medical practice that involves ta… #
Related terms include personalized medicine and targeted therapy. In healthcare, precision medicine can be applied to cancer treatment and genetic disorders.
Principal Component Analysis (PCA) is a technique used to reduce t… #
Related terms include t-distributed Stochastic Neighbor Embedding and autoencoders. In healthcare, PCA can be applied to genomic data to identify genetic variants associated with diseases.
Random Forest is a type of machine learning algorithm that uses multiple… #
Related terms include gradient boosting and support vector machines. In healthcare, random forest can be applied to clinical decision support systems to predict patient outcomes and recommend treatments.
Recurrent Neural Networks (RNNs) are a type of deep learning model that u… #
Related terms include long short-term memory and gated recurrent units. In healthcare, RNNs can be applied to clinical decision support systems to predict patient outcomes and recommend treatments.
Regression is a type of machine learning task that involves predicting… #
Related terms include linear regression and logistic regression. In healthcare, regression can be applied to clinical decision support systems to predict patient outcomes and recommend treatments.
Regularization is a technique used to prevent overfitting i… #
Related terms include dropout and early stopping. In healthcare, regularization can be applied to deep learning models for image classification tasks, such as tumor detection in medical images.
Risk Stratification is the process of identifying high #
risk patients or populations based on their clinical characteristics and genetic profiles, often used in healthcare to target interventions and improve patient outcomes. Related terms include predictive analytics and precision medicine. In healthcare, risk stratification can be applied to population health management and clinical decision support.
Semi #
Supervised Learning is a type of machine learning that uses a combination of labeled and unlabeled data to train models, often used in healthcare to improve the accuracy of models and reduce labeling costs. Related terms include active learning and transfer learning. In healthcare, semi-supervised learning can be applied to image classification tasks, such as tumor detection in medical images.
Sensitivity is a measure of a test or model 's ability to detect… #
Related terms include specificity and accuracy. In healthcare, sensitivity is crucial for diagnostic models and treatment recommendations.
Specificity is a measure of a test or model 's ability to detect… #
Related terms include sensitivity and accuracy. In healthcare, specificity is crucial for diagnostic models and treatment recommendations.
Support Vector Machines (SVMs) are a type of machine learning algorithm t… #
Related terms include random forest and gradient boosting. In healthcare, SVMs can be applied to clinical decision support systems to predict patient outcomes and recommend treatments.
Supervised Learning is a type of machine learning that involves training… #
Related terms include unsupervised learning and semi-supervised learning. In healthcare, supervised learning can be applied to clinical decision support systems to predict patient outcomes and recommend treatments.
Targeted Therapy is a type of cancer treatment that involves targeting… #
Related terms include precision medicine and personalized medicine. In healthcare, targeted therapy can be applied to cancer treatment and genetic disorders.
Transfer Learning is a type of machine learning that involves transfer… #
Related terms include active learning and semi-supervised learning. In healthcare, transfer learning can be applied to image classification tasks, such as tumor detection in medical images.
Unsupervised Learning is a type of machine learning that involves disc… #
Related terms include clustering and dimensionality reduction. In healthcare, unsupervised learning can be applied to genomic data to identify genetic variants associated with diseases.
Validation is the process of evaluating the performance of a ma… #
Related terms include cross-validation and bootstrapping. In healthcare, validation is crucial for model deployment and clinical decision-making.