Technically Speaking

The Official Bigstep Blog

 

Data Science Webinar: Recommender Systems - From Simple to Complex

On November 15th, Bigstep will host a webinar on the topic of Recommender Systems, ranging from simple systems to complex ones. Registration is limited, so save your spot now.  The webinar will be hosted by Andras Palfi, Data Scientist at Bigstep, who recently gave a talk at Big Data Week London Conference on Automation in Data Science and Machine Learning.

On November 15th, Bigstep will host a webinar on the topic of Recommender Systems, ranging from simple systems to complex ones. Registration is limited, so save your spot now. The webinar will be hosted by Andras Palfi, Data Scientist at Bigstep, who recently gave a talk at Big Data Week London Conference on Automation in Data Science and Machine Learning.

Recommender systems are a collection of algorithms that can be used to personalize content and offers for customers. Often considered one of the most successful and widespread application of machine learning technologies in business, they have been employed by major companies like Netflix, Amazon or Google to create new revenue streams and provide tailored experiences. Fortunately, their benefits are not limited to major tech companies with deep pockets. With a minimal technical background, almost anyone can implement a simple recommender system.>

The objective of this webinar is to demystify the algorithms and methods used by recommender systems and provide a few practical use cases. No previous technical background is necessary, but familiarity with high-school level programming and mathematics will be helpful.

There are two main methods to construct a recommender system:

1. Content-based recommendation

2. Collaborative Filtering

The first one analyzes the nature of each item. For instance, recommending books to a user by performing Natural Language Processing on the content of each book.

Collaborative Filtering, on the other hand, does not require any information about the items or the users themselves. It recommends items based on users’ past behavior. Collaborative Filtering methods can be broken down into two categories:

1. Memory-based approach

2. Model-based approach

Memory-based approaches make recommendations by looking at similarities between items or similarities between users. The key idea behind is that similar users share the same interest and that similar items are liked by a user. On the other hand, model-based approaches use various latent factor models to capture the similarity between users and items. Essentially, they turn the recommendation problem into an optimization problem.

The webinar will focus on implementing both a memory-based and a model-based recommendation system in Python.

Join Andras Palfi in this interactive webinar on November 15. See on our blog an example of how Nexflix implemented a recommendation algorithm.

Got a question? Need advice? We're just one click away.
Sharing is caring:TwitterFacebookLinkedinPinterestEmail

Readers also enjoyed:

'Data Scientist' Earns Top Spot in 'Best Jobs in America'

What do you think the best job in the country is? One guy works as a personal shopper for clients looking for exotic sports cars--that's a cool job.Another…

Leave a Reply

Your email address will not be published.

* Required fields to post your comments.
Please review our Privacy Notice in order to understand how we process your personal data and what are your rights in this respect.