Skip to main content

Concepts

In the following sections, we will more systematically introduce the following concepts:

📄️ Bias & Fairness

It can’t be denied that there is bias all around us. A bias is a prejudice against a person or group of people, including, but not limited to their gender, race, and beliefs. Many of these biases arise from emergent behavior in social interactions, events in history, and cultural and political views around the world. These biases affect the data that we collect. Because AI algorithms work with this data, it is an inherent problem that the machine will “learn” these biases. From a technical perspective, we can engineer the system perfectly, but at the end of the day, humans interact with these systems, and it’s our responsibility to minimize bias and prejudice as much as possible. The algorithms we use are only as good as the data provided to them. Understanding the data and the context in which it is being used is the first step in battling bias, and this understanding will help you build better solutions—because you will be well versed in the problem space. Providing balanced data with as little bias as possible should result in better solutions.

📄️ Cross-domain

A common challenge for most current recommender systems is the cold-start problem. Due to the lack of user-item interactions, the fine-tuned recommender systems are unable to handle situations with new users or new items. Recently, some works introduce the meta-optimization idea into the recommendation scenarios, i.e. predicting the user preference by only a few of past interacted items. The core idea is learning a global sharing initialization parameter for all users and then learning the local parameters for each user separately. However, most meta-learning based recommendation approaches adopt model-agnostic meta-learning for parameter initialization, where the global sharing parameter may lead the model into local optima for some users.

📄️ Data Science

Data science is used in a variety of ways. Some data scientists focus on the analytics side of things, pulling out hidden patterns and insights from data, then communicating these results with visualizations and statistics. Others work on creating predictive models in order to predict future events, such as predicting whether someone will put solar panels on their house. Yet others work on models for classification; for example, classifying the make and model of a car in an image. One thing ties all applications of data science together: the data. Anywhere you have enough data, you can use data science to accomplish things that seem like magic to the casual observer.

📄️ Incremental Learning

Humans and animals have the ability to continually acquire, fine-tune, and transfer knowledge and skills throughout their lifespan. This ability, referred to as lifelong learning, is mediated by a rich set of neurocognitive mechanisms that together contribute to the development and specialization of our sensorimotor skills as well as to long-term memory consolidation and retrieval. Consequently, lifelong learning capabilities are crucial for autonomous agents interacting in the real world and processing continuous streams of information. However, lifelong learning remains a long-standing challenge for machine learning and neural network models since the continual acquisition of incrementally available information from non-stationary data distributions generally leads to catastrophic forgetting or interference.

📄️ MLOps

The boom in AI has seen a rising demand for better AI infrastructure — both in the compute hardware layer and AI framework optimizations that make optimal use of accelerated compute. Unfortunately, organizations often overlook the critical importance of a middle tier: infrastructure software that standardizes the ML life cycle, adding a common platform for teams of data scientists and researchers to standardize their approach and eliminate distracting DevOps work. This process of building the ML life cycle is increasingly known as MLOps, with end-to-end platforms being built to automate and standardize repeatable manual processes. Although dozens of MLOps platforms exist, adopting one can be confusing and cumbersome. What should be considered when employing MLOps? What are the core pillars to MLOps, and which features are most critical?

📄️ Multi-Objective Optimization

Recommender systems have been widely applied to several domains and applications. Traditional recommender systems usually deal with a single objective, such as minimizing the prediction errors or maximizing the ranking of the recommendation list. There is an emerging demand for multi-objective optimization so that the development of recommendation models can take multiple objectives into consideration, especially in the area of multi-stakeholder and multi-task recommender systems.