In many real-world problems, unlabeled data are often abundant whereas labeled data are scarce. Label acquisition is usually expensive due to the involvement of human experts, and thus, it is important to train an accurate prediction model by a small number of labeled instances. Active learning aims at reducing human efforts on annotating examples in a machine learning system, and has been successfully applied into various real tasks.
Brain imaging technology is one of the important tools for brain science studies. Brain imaging data itself has characteristics of high dimensionality, heterogeneity and time-varying. How to analyze the brain imaging data quickly and effectively is one of the key issues in current research. At the same time, the brain has a high degree of complexity in structure and function, and traditional brain image analysis methods are difficult to reveal complex and subtle patterns of brain changes.
Collecting groundtruth labels in many tasks are challenging or impractical. By distributing the task to multiple workers and estimating the true labels via some aggregation schemes, crowdsourcing provides a new learning paradigm, and has achieved applications in various fields such as sentiment classification, medical diagnosis and image tagging.
Computer vision is a science that studies how to make the machine see. Furthermore, it refers to the use of camera and computer instead of human eye to recognize, track and measure the target, and further do graphics processing, so that computer processing becomes more suitable for human eye observation or transmitted to the instrument for detection. As a scientific discipline, computer vision studies related theories and technologies, trying to establish an artificial intelligence system.
Face forgery recognition is aimed at the current popular face forgery technology, such as Deepfake. The main work includes: generating forgeries, detecting forgeries (if local forgeries, marking and classifying the forgeries; if global forgeries, classifying the forgeries), and evaluating the identification model. The main methods include Face X-ray, Style-Gan and so on. Face forgery recognition plays an important role in user authentication, privacy protection and other fields.
Gene expression level is how much each gene is quantified in biological individuals. The differential expression analysis of gene can essentially reveal the pathogenic mechanism behind the diseased individual. Gene expression analysis is to use data obtained from different sequencing technologies to model and analyze gene expression level and differentially expressed genes.
Numerical simulation is a kind of technology relied on computers to solve all kinds of problems in nature and engineering through modeling, finite element analysis, numerical calculation and image display. Machine learning is widely utilized to assist some stages of numerical simulation so as to efficiently obtain more accurate simulation results.
As a new distributed computing technology, multi-agent system has become an idea method and tool for complex system analysis and simulation. A multi-agent system is a computing system composed of multiple agents interacting in one environment. Multi-agent system can also be used to solve the problems that are difficult to be solved by separate agents and single-layer systems. Intelligence can be achieved by methods, functions, processes, search algorithms, or reinforcement learning.
Reinforcement learning is one of the paradigms and methodologies of machine learning, which is used to describe and solve the problem that agents maximize returns or achieve specific goals through learning strategies in the process of interaction with the environment. Inspired by behaviorism psychology, reinforcement learning focuses on online learning and tries to maintain a balance between exploration-exploitation.