Book Image

Building a Recommendation System with R

Book Image

Building a Recommendation System with R

Overview of this book

Table of Contents (13 chapters)
Building a Recommendation System with R
Credits
About the Authors
About the Reviewer
www.PacktPub.com
Preface
References
Index

Index

A

  • accuracy metrics
    • precision / Evaluating the recommendations
    • recall / Evaluating the recommendations
  • Akaike information criteria (AIC) / Evaluating data-mining algorithms
  • area under the curve (AUC)
    • about / Identifying the most suitable model

B

  • bagging (Bootstrap aggregating) / Bagging
  • Bayesian information criteria (BIC) / Evaluating data-mining algorithms
  • binary data
    • collaborative filtering / Collaborative filtering on binary data
    • item-based collaborative filtering / Item-based collaborative filtering on binary data
    • user-based collaborative filtering / User-based collaborative filtering on binary data

C

  • case study
    • about / A case study
  • cluster analysis
    • about / Cluster analysis
    • K-means cluster algorithm / Explaining the k-means cluster algorithm
  • collaborative filtering
    • conclusions / Conclusions about collaborative filtering
    • limitations / Limitations of collaborative filtering
  • collaborative filtering recommender systems
    • about / Collaborative filtering recommender systems
  • Comprehensive R Archive Network (CRAN) / Decision trees
  • confusion matrix / Evaluating data-mining algorithms
  • content-based filtering
    • about / Content-based filtering
  • content-based recommender systems
    • about / Content-based recommender systems
  • cosine distance / Cosine distance

D

  • data
    • preparing / Preparing the data
    • describing / Description of the data
    • record, types / Description of the data
    • importing / Importing the data
    • rating matrix, defining / Defining a rating matrix
    • item attributes, extracting / Extracting item attributes
  • data-mining algorithms
    • evaluating / Evaluating data-mining algorithms
    • cross validation / Evaluating data-mining algorithms
    • regularization / Evaluating data-mining algorithms
    • confusion matrix / Evaluating data-mining algorithms
    • precision / Evaluating data-mining algorithms
    • Recall/Sensitivity / Evaluating data-mining algorithms
    • specificity / Evaluating data-mining algorithms
    • model comparison / Evaluating data-mining algorithms
  • data analysis problem
    • solving / Solving a data analysis problem
  • data exploration
    • about / Data exploration
    • nature of data, exploring / Exploring the nature of the data
    • values of rating, exploring / Exploring the values of the rating
    • viewed movies, exploring / Exploring which movies have been viewed
    • average ratings, exploring / Exploring the average ratings
    • matrix, visualizing / Visualizing the matrix
  • data mining
    • techniques / Data mining techniques
  • data preparation
    • about / Data preparation
    • most relevant data, selecting / Selecting the most relevant data
    • most relevant data, exploring / Exploring the most relevant data
    • data, normalizing / Normalizing the data
    • data, binarizing / Binarizing the data
    • for evaluating models / Preparing the data to evaluate the models
    • data, splitting / Splitting the data
    • data, bootstrapping / Bootstrapping data
    • k-fold, using / Using k-fold to validate models
  • data preprocessing techniques
    • about / Data preprocessing techniques
    • similarity measures / Similarity measures
    • dimensionality reduction / Dimensionality reduction
  • datasets
    • training set / Preparing the data to evaluate the models
    • testing set / Preparing the data to evaluate the models
  • decision trees
    • about / Decision trees
    • using / Decision trees
  • dimensionality reduction
    • about / Dimensionality reduction
    • Principal component analysis (PCA) / Principal component analysis

E

  • ensemble methods
    • about / Ensemble methods
    • bagging (Bootstrap aggregating) / Bagging
    • random forests / Random forests
    • boosting / Boosting
  • Euclidian distance / Euclidian distance
  • evaluation techniques
    • about / Evaluation techniques

F

  • False Positive Rate (FPR)
    • about / Evaluating the recommendations
  • function
    • defining / Building a function to evaluate the model
    • data / Building a function to evaluate the model
    • k-fold parameters / Building a function to evaluate the model
    • model parameters / Building a function to evaluate the model

H

  • hybrid recommender systems
    • about / Hybrid recommender systems
  • hybrid systems
    • about / Hybrid systems

I

  • item-based collaborative filtering
    • about / Item-based collaborative filtering
    • test set, defining / Defining the training and test sets
    • training set, defining / Defining the training and test sets
    • recommendation model, building / Building the recommendation model
    • recommendation model, exploring / Exploring the recommender model
    • recommender model, applying on test set / Applying the recommender model on the test set
    • on binary data / Item-based collaborative filtering on binary data

K

  • k-fold
    • used, for validating models / Using k-fold to validate models
  • K-means cluster algorithm
    • about / Explaining the k-means cluster algorithm
    • cluster assignment step / Explaining the k-means cluster algorithm
    • move centroid step / Explaining the k-means cluster algorithm
    • support vector machines (SVM) / Support vector machine
  • knowledge-based recommender systems
    • about / Knowledge-based recommender systems, Knowledge-based recommender systems

M

  • mean absolute error (MAE)
    • about / Evaluating the ratings
  • mean squared error (MSE)
    • about / Evaluating the ratings
  • monolithic hybrid systems
    • feature combination / Hybrid recommender systems
    • feature augmentation / Hybrid recommender systems

P

  • Pearson correlation / Pearson correlation
  • Principal component analysis (PCA) / Data preprocessing techniques, Principal component analysis
  • proper model
    • identifying / Identifying the most suitable model
    • models, comparing / Comparing models
    • most proper model, identifying / Identifying the most suitable model
    • numeric parameter, optimizing / Optimizing a numeric parameter

R

  • recommendation model
    • building / Building the model
    • optimizing / Evaluating and optimizing the model
    • evaluating / Evaluating and optimizing the model
    • evaluating, by building function / Building a function to evaluate the model
    • parameters, optimizing / Optimizing the model parameters
  • recommenderlab
    • used, for building recommender systems / R package for recommendation – recommenderlab
    • datasets / Datasets
    • MovieLense dataset / Jester5k, MSWeb, and MovieLense
    • MSWeb dataset / Jester5k, MSWeb, and MovieLense
    • Jester5k dataset / Jester5k, MSWeb, and MovieLense
    • realRatingMatrix class, defining / The class for rating matrices
    • similarity matrix, computing / Computing the similarity matrix
    • recommendation models, displaying / Recommendation models
  • recommender systems
    • about / Understanding recommender systems
    • case study / A case study
    • future scope / The future scope
    • building, recommenderlab used / R package for recommendation – recommenderlab
  • recommender techniques
    • evaluating / Evaluating recommender techniques
    • ratings, evaluating / Evaluating the ratings
    • recommendations, evaluating / Evaluating the recommendations
  • root mean square error (RMSE)
    • about / Evaluating the ratings
  • R package
    • recommenderlab / R package for recommendation – recommenderlab

S

  • similarity measures
    • about / Similarity measures
    • Euclidian distance / Euclidian distance
    • Cosine distance / Cosine distance
    • Pearson correlation / Pearson correlation
  • singular value decomposition (SVD) / Data preprocessing techniques
  • support vector machines (SVM) / Support vector machine

T

  • test set
    • about / Defining the training and test sets
  • training set
    • about / Defining the training and test sets
  • True Positive Rate (TPR)
    • about / Evaluating the recommendations

U

  • user-based collaborative filtering
    • about / User-based collaborative filtering
    • recommendation model, building / Building the recommendation model
    • recommender model, applying on test set / Applying the recommender model on the test set
    • on binary data / Collaborative filtering on binary data, User-based collaborative filtering on binary data
    • data preparation / Data preparation