Εξόρυξη γνώσης από Βάσεις Δεδομένων και τον Παγκόσμιο Ιστό - Open Courses (INF108)

Εισαγωγή στις μεθόδους Dimesionality Reduction and Feature Selection, Supervised learning, Unsupervised Learning (Clustering), Community Detection and Evaluation in Social and Information Networks, Web Mining, Introduction to Big Data που θα δούμε πιο αναλυτικά στις επόμενες ενότητες.

Εισαγωγή και εξοικείωση με τις μεθόδους Pre-processing, Exploration, Feature selection, Dimensionality reduction, feature extraction and evaluation.

Λέξεις κλειδιά: Distance Measures, Exploratory Data Analysis, Data Quality, Evaluation metrics, Filter methods, Wrapper methods, Embedded methods, Complexity, Data features, k-NN, Linear Algorithms, Matrix Factorization methods, Principal Components Analysis (PCA) , Singular Value Decomposition (SVD), Multidimensional Scaling (MDS), Non negative, Matrix Factorization (NMF), Latent Semantic Indexing (LSI)

Εισαγωγή και εξοικείωση με τις μεθόδους k-nn, regression, logistic regression, decision trees.

Λέξεις κλειδιά: Prediction, Regression, Least Squares method, Bayesian classification, Bayes theorem, Naïve Bayes Classifiers, Linear Discriminant Analysis, k-NN, Boundary, Decision Tree, Entropy, Information gain, Learning – linear regression, LMS

Εισαγωγή και εξοικείωση με τις μεθόδους, Clustering, K-means, Expectation Maximization (EM), Spectral Clustering.  

Λέξεις κλειδιά: 

Clustering, Partitional method, K-means, time complexity,  speed-up tricks, Expectation Maximization, Gaussian Mixture Models, E step, M step, Hierarchical Clustering, Association Mining, Apriori Algorithm, Spectral Clustering, Graph-Cut, Min-Cut, Graph Laplacian, Ratio Cut, Normalized Cut, Algorithm for k>2, Laplacian-Eigenvectors-EigenValues, Modularity function, Newman-Girvan algorithm

Εισαγωγή και εξοικείωση με τις μεθόδους Introduction & Motivation, Community evaluation measures, Graph clustering algorithms, Alternative Methods for Community Evaluation, New directions for research in the area of graph mining.

Λέξεις κλειδιά: Networks, Social Networks, Community Detection, World Wide Web, Metabolic networks, intra-clusters,  inter-clusters, Notation, Evaluation based on internal connectivity, Evaluation based on external connectivity, Evaluation based on network model, Observation, Graph-Cut, Min-Cut, Graph Laplacian, Ratio Cut, Normalized Cut, Modularity, Newman-Girvan algorithm, optimization, Resolution Limit, conductance, Network Community Profile (NCP), Core-periphery, Jellyfish model, Robustness, K-core, DBLP K-cores, Wikipedia, S-cores, CoreCluster framework, Scalability

Εισαγωγή και εξοικείωση με τις μεθόδους Web personalization and recommendations (collaborative filtering), Web Advertising.

Λέξεις κλειδιά: Personalization, recommendations, Recommender Data Model, Scalability, Privacy and Trust, Robustness, Utilization, User Interests, Collaborative Filtering (CF), User-User Similarity, CF Recommender, Item-to-Item Collaborative filtering ,  Advertising, Traditional advertising,  Online advertising, Cost Per Impression (CPI), Cost Per Click (CPC) ,  Cost Per Action (CPA), Google AdWords

Εισαγωγή και εξοικείωση με τις μεθόδους Map reduce - distributed processing, technologies (Hadoop, Map Reduce,  NoSQL storage)

Λέξεις κλειδιά: Big Data, Hadoop, HDFS, MapReduce, Pig, Hive, HBase, MapReduce, Map function, Data Flow, Map task, HDFS, Hive, HiveQL, Querying Data

-