
Chapter 1. Health Data Management
a. Introduction to the fundamentals of Data Science.
b. Health data types (vital signs, ΜRI, ECG, CT scans/images, waveforms/signals, etc.)
c. Electronic Health Record (EHR) (Demographics, Diagnoses, Medical Interventions, Prescriptions, Medications/Drugs, Side effects, Lab/Microbiology Tests, Progress notes)
d. MIMIC Data set
i. Description
ii. Data Types
iii. Data Preprocessing (missing values, etc.)
iv. Data base schema
e. Programming Exercise with python for Analysis and visualization of medical data using
libraries such as Pandas, NumPy, Matplotlib, etc
10 SAMPLE Slides

Chapter 2. Privacy Protection of Sensitive Medical Data
a. Anonymization Techniques
b. Date shifting
c. Format Conversion
d. Generalization of data interactions
e. KAnonymity
f. Adding Noise in Data
e. Loss of Information
g. Differential Privacy (Central and Local DP) for sensitive health data.
8 SAMPLE Slides
Python code in Google Colab

Chapter 3. Classification algorithms for Static Data and Time Series.
a. Linear Regression
b. Logistic Regression
c. Decision Tree
d. Random Forest
e. Naïve Bayes Classifier
f. Feature Selection
i. Gini index
ii. Entropy
iii. x^2 statistic
.
Python code in Google Colab

Chapter 4. Clustering of Medical Data and Genetic Algorithms.
a. Clustering Algorithms
i. Kmeans
ii. Hierarchical Clustering
iii. DBSCAN
iv. BFR algorithm (Bradley, Fayyad, and Reina)
b. Genetic Algorithms
i. Genetic Operations
ii. The Architecture of a Genetic Algorithm
iii. The Genetic Algorithm in pseudocode form
iv. StepbyStep Execution of the Genetic Algorithm
c. Programming Exercise with python for clustering with kmeans algorithm the patients into groups of diabetes disease or not
10 SAMPLE Slides
Python code in Google Colab

Chapter 5. Matrix Factorization for Health Data
a. PCA decomposition
b. Singular Value Decomposition
c. UVdecomposition
d. CURdecomposition
e. Tensor Decomposition
f. Programming Exercise with python to apply PCA to perform dimensionality reduction and bring into surface latent associations between drugs and unwanted side effects.
Python code in Google Colab

Chapter 6. Machine learning algorithms for Image Processing and other complex medical signals
a. Convolution Neural Networks
I. CNN architecture
II. Pooling Layers
III. ResNet
b. Applications over different medical signals
i. Electrocardiogram ECG
ii. Magnetic Resonance Imaging (MRI)
iii. Computed Tomography Scan
c. Support Vector machines
I. Finding Optimal Separators with Gradient Descent
II. Hard and Soft SVMs
d. Programming Exercise with python to apply image processing over CT scans and classify patients into COVID and nonCOVID disease.
9 SAMPLE Slides
Python code in Google Colab

Chapter 7. Reinforcement Learning, and Deep Neural Nets
a. Markov Chain
b. Qlearning Algorithm
c. Deep Reinforcement Learning
i. Deep QNetwork
ii. Double Deep QNetwork/A2C
iii. Optimal Insulin Dose Prediction for Diabetes Patients
c. Multilayer perceptron
i. Activation Functions
ii. Loss Functions
iii. Regression Loss
iv. Classification Loss
d. Recurrent Neural Networks
i. Vanishing and Exploding Gradients
ii. Long Sort Term Memory LSTM
e. Programming Exercise with Python to predict the optimal insulin dose for patient with diabetes using tabular Qlearning algorithm.
<10 SAMPLE Slides
Python code in Google Colab

Chapter 8. Graph algorithms
a. Local based similarity algorithms (Shortest Path
i. Common Neighbors, Jaccard similarity index, Salton similarity index, Adamic & Adar similarity index, Preferential Attachement.
b. Globalbased algorithms
i. Random Walk with Restart
ii. SimRank
iii. PathSim
c. Graph Convolution networks
d. Graph Embeddings
e. Programming Exercise with Python to predict drug combinations to patients and explain the predictions using graph data (patient nodes, treatment node, drug node).
10 SAMPLE Slides
Python code in Google Colab

Chapter 9. Evaluation Metrics of Prediction Models
a. Confusion Matrix
b. Precision, Recall, and NDCG
c. PrecisionRecall and ROC curves
d. MAE, RMSE for Optimal Drug Dose Prediction
e. Beyond Accuracy Metrics
i. Explainability in Health
f. Programming Exercise with Python to recommend drug combinations to patients and measure quantitatively the predictions along with the unwanted side effects they have.
Python code in Google Colab

Chapter 10. New Trends and Apps in Health Care
i. Smart Medical Watches
ii. Smart Medical Devices.
b. Large Language Models, and Prompt Engineering
i. Application of AI algorithms over medical progress notes.
ii. Vector space model and TFIDF
iii ChatGPT and RAG
c. Programming Exercise with Python to send the electronic health record of a patient to ChatGPT API using a prompt request (prompt engineering) and get a possible therapeutic prediction based on the given vital signs.
10 SAMPLE Slides
