APPLIED MACHINE LEARNING LAB

Artificial Intelligence (AI) includes the following three perspectives*:

1. The classical “human-imitative” perspective, such as AI in the movies, interactive home robotics

2. The “intelligence augmentation” perspective

   • The system need not be intelligent itself, but it reveals patterns that humans can make use of

   • cf. search engines, recommendation systems, natural language translation, predictive maintenance

3. The “intelligent infrastructure” perspective

   • where large-scale, distributed collections of dataflows and loosely-coupled decisions

   • cf. smart city, smart building, digital farming, smart manufactory, digital logistic

The VISION of Applied Machine Learning Lab is that our world is moving toward a state where the physical world is captured by sensors and is encapsulated in a digital model in which human-centric machine learning (ML) is essential to useful, trustworthy, and responsible ambient intelligence.

Our MISSION is to research and develop such an ambient intelligence infrastructure** and its applications as well as resolve issues associated with the “augmented Intelligence” perspective, such as:

• Human-centric ML in which data-driven ML is augmented with human feedback to achieve responsible ML

• Such augmentation can be verified for fairness, privacy, and robustness as well as required enforcement by governmental laws and policy, corporate strategies, and ethical principles.

• Support interaction/fusion between ML and human, such as interpretation, augmentation, verification, influence, integration, and causal inference

• Derive best practices and principles for continuous software development and implementation cycle where ML is a crucial component of the software decision-making functionalities.