With the progress in time, we do not have to rely on crystal balls any more to predict the future, we have data! Recommender systems or Recommendation Engines serve as the modern-day crystal balls, with the exception that all of the predictions made by them are backed by data!
Recommender Systems also perform the task of filtering, prioritizing and efficiently delivering relevant information in order to alleviate the problem of information overload, which has created a potential problem to many users.
With all these advantages, Recommendation Engines are very common these days and can be applied in almost every field.
Packt’s Video Learning Paths are an amalgamation of multiple video courses that are logically tied together to provide you with a larger learning curve.
In this Learning Path, you will be introduced to what a recommendation engine is and its applications. You will then learn to build recommender systems by using popular frameworks such as R and Python.
The latter part of the Learning Path will deal with various complex recommendation engines such as personalized recommendation engines, real-time recommendation engines, and SVD recommender systems. You will also get a quick glance into the future of recommendation systems.
By the end of this Learning Path, you will be able to build efficient recommendation engines by following the best practices.
For this Learning Path, we have taken two video courses both authored by Suresh Kumar Gorakala.
Suresh Kumar Gorakala is a Data scientist focused on Artificial Intelligence. He has professional experience close to 10 years, having worked with various global clients across multiple domains and helped them in solving their business problems using Advanced Big Data Analytics. He has extensively worked on Recommendation Engines, Natural language Processing, Advanced Machine Learning, and Graph Databases. He previously co-authored Building a Recommendation System with R for Packt Publishing. He is a passionate traveler and is a photographer by hobby.
Learn to provide relevant suggestions based on datamining approaches
know about the different types of popular recommender systems and its use
Learn to use the large user base and products which can scale easily and respond fast
Learn to install R Package in Rstudio to know how to load and format data
Learn to use the correlation value as the measure of similarity between two items in a matrix.
Learn to predict the unrated movies of a specified user using the ratings given by similar users
Learn to develop new and different approaches
Learn to extract the features that represent the product
Learn to capture the context information of the user and refine their suggestions accordingly.
Learn to implement the hybrid recommendation engine based on the problem statement and business needs
Learn to build a probability modelusing the prior probabilities from the available data
Learn to know the simple mathematical calculation that is applied between two vectors
Know about matrix factorization, alternate least squares, and singular value decomposition
Learn to predict the future outcomes based on given input parameters
Learn linear classification, KNN classification, support vector machines, decision trees, and various ensemble methods
Learn to know the process of grouping objects in one Group
Learn to compute the similarity measure metric that data should be all numeric
Learn about cross validation and various popular evaluation metrics like root mean square error
The recommenderlab R package is a framework for developing and testing recommendation algorithms used to build recommendation engines. In this video, we’ll see how to installrecommenderlab
Like any other package available in R, recommenderlab also comes with default datasets. We need to know how to see available packages, methods, and algorithms.
Now that we have started with the basics of data set, we must explore the data in more detail. This video will do exactly that and guide us in building and evaluating a recommender model
We use the same Jester5k dataset for the item-based recommender system as with UBCF. But, removing data of certain users and building IBCF system can be a bit tricky. Let’s see how we do that
Now, we have seen implementations of user-based recommender systems and item-based recommender systems using the R package, recommenderlab. But, can we do the same with Python? In this video, we see theUBCF and IBCF implementations using the Python programming language
Let’s explore the MovieLens dataset and also prepare the data required for building collaborative filtering recommendation engines using python
If we observe the RMSE values in the model, we can see that the error is a bit higher. The reason may be that we have chosen all the users' rating information while making the predictions. In this video, we will cover finding the top-N nearest neighbors.
IBCF is very similar to UBCF but with very minor changes in how we use the rating matrix. But, still we must learn how to evaluate the model. This video guides in evaluating the model.
Learn to know the flavors of personalized recommenders.
Build content recommendations using another approach, using the Python sklearn, NumPy, and pandas packages.
Use different recommendations to the same person based on their current context.
Create context profile of the user.
Learn to provide the capabilities of Spark, such as in-memory distributed computation and fast, easy-to-use APIs.
Learn to build a specific version of Hadoop to access HDFS as well as standard and custom Hadoop input sources.
Know the Matrix Factorization Model and the Alternating Least Squares method.
Learn to build the recommendation engine using Sparksuch as DataFrames, RDD, Pipelines, and Transforms available in Spark MLlib.
Know the actual implementation of the recommendation engine.
Learn to choose the Root Mean Squared Error method to calculate the model accuracy.
Learn to understand the concept of databases and where to apply them.
Learn the cypher query language.
Learn to create nodes and relationships.
Learn how to install Neo4j for Windows.
Learn to download and install Neo4j on the CentOS Linux Platform.
Build recommendation engines using the interface.
Learn to write a query to generate recommendation.
Ability to implement collaborative filtering using Euclidean distance method.
Ability to implement collaborative filtering using cosine similarity.
Learn to setup the Apache mahout software.
Build customized recommender systems that are enterprise-ready, scalable, flexible, and that perform well.
Item-based recommenders recommend similar items to users by considering the similarity between items instead of the similarity between users.
Evaluate the accuracy of the recommender models that we built.
Use of matrix factorization methods to generate model-based recommender implementations in Mahout.
Understand the directions in which recommendation engines are evolving to cope with futuristic situations.
List a few promising use cases that might make you more interested in future of recommendation engines.
Learn to build recommendation engines for improving the robustness and relevance of the recommendations.
Packt has been committed to developer learning since 2004. A lot has changed in software since then - but Packt has remained responsive to these changes, continuing to look forward at the trends and tools defining the way we work and live. And how to put them to work.
With an extensive library of content - more than 4000 books and video courses -Packt's mission is to help developers stay relevant in a rapidly changing world. From new web frameworks and programming languages, to cutting edge data analytics, and DevOps, Packt takes software professionals in every field to what's important to them now.
From skills that will help you to develop and future proof your career to immediate solutions to every day tech challenges, Packt is a go-to resource to make you a better, smarter developer.
Packt Udemy courses continue this tradition, bringing you comprehensive yet concise video courses straight from the experts.