Idiom General (Articles)

The future of AI

The future of AI

If you wonder what is next in the evolution towards general AI then this session is for you. We have seen some painful failures of artificial intelligence pointing to a lack of ‘common sense’. Are neural networks really the solution we seek or is a new path needed? Find out what IBM Research is cooking in terms of hardware and software in the never ending quest towards General AI.

The future of AI Leer más »

IEEE Use Case–Criteria for Addressing Ethical Challenges in Transparency, Accountability, and Privacy of CTA/CTT

IEEE Use Case–Criteria for Addressing Ethical Challenges in Transparency, Accountability, and Privacy of CTA/CTT

There are substantial public health benefits gained through successfully alerting individuals and relevant public health institutions of a person’s exposure to a communicable disease. Contact tracing techniques have been applied to epidemiology for centuries, traditionally involving a manual process of interview and follow-up. This is time-consuming, difficult, and dangerous work. Manual processes are also open to incomplete information because they rely on individuals being willing and able to remember and report all contact possibilities.

IEEE Use Case–Criteria for Addressing Ethical Challenges in Transparency, Accountability, and Privacy of CTA/CTT Leer más »

The Deep Learning Revolution and Its Implications for Computer Architecture and Chip Design

The Deep Learning Revolution and Its Implications for Computer Architecture and Chip Design

The past decade has seen a remarkable series of advances in machine learning, and in particular deep learning approaches based on artificial neural networks, to improve our abilities to build more accurate systems across a broad range of areas, including computer vision, speech recognition, language translation, and natural language understanding tasks.

The Deep Learning Revolution and Its Implications for Computer Architecture and Chip Design Leer más »

TextAttack: A Framework for Adversarial Attacks, Data Augmentation, and Adversarial Training in NLP

TextAttack: A Framework for Adversarial Attacks, Data Augmentation, and Adversarial Training in NLP

This paper introduces TextAttack, a Python framework for adversarial attacks, data augmentation, and adversarial training in NLP. TextAttack builds attacks from four components: a goal function, a set of constraints, a transformation, and a search method. TextAttack’s modular design enables researchers to easily construct attacks from combinations of novel and existing components. TextAttack provides implementations of 16 adversarial attacks from the literature and supports a variety of models and datasets, including BERT and other transformers, and all GLUE tasks.

TextAttack: A Framework for Adversarial Attacks, Data Augmentation, and Adversarial Training in NLP Leer más »

Diagnostic uncertainty calibration: towards reliable machine predictions in medical domain

Diagnostic uncertainty calibration: towards reliable machine predictions in medical domain

We further formalize the metrics for higher-order statistics, including inter-rater disagreement, in a unified way, which enables us to assess the quality of distributional uncertainty. In addition, we propose a novel post-hoc calibration method that equips trained neural networks with calibrated distributions over class probability estimates. With a large-scale medical imaging application, we show that our approach significantly improves the quality of uncertainty estimates in multiple metrics.

Diagnostic uncertainty calibration: towards reliable machine predictions in medical domain Leer más »