choi


New Model Allows Earlier, More Accurate Air Pollution Warnings

#artificialintelligence

That would improve health alerts for people at heightened risk of developing problems because of high ozone levels. Yunsoo Choi, associate professor in the Department of Earth and Atmospheric Sciences and corresponding author for a paper explaining the work, said they built an artificially intelligent model using a convolutional neural network, which is able to take information from current conditions and accurately predict ozone levels for the next day. The work was published in the journal Neural Networks. "If we know the conditions of today, we can predict the conditions of tomorrow," Choi said. Ozone is an unstable gas, formed by a chemical reaction when sunlight combines with nitrogen oxides (NOx) and volatile organic compounds, both of which are found in automobile and industrial emissions.


Deep learning assists in detecting malignant lung cancers

#artificialintelligence

Radiologists assisted by deep-learning based software were better able to detect malignant lung cancers on chest X-rays, according to research published in the journal Radiology. "The average sensitivity of radiologists was improved by 5.2% when they re-reviewed X-rays with the deep-learning software," said Byoung Wook Choi, M.D., Ph.D., professor at Yonsei University College of Medicine, and cardiothoracic radiologist in the Department of Radiology in the Yonsei University Health System in Seoul, Korea. "At the same time, the number of false-positive findings per image was reduced." Dr. Choi said the characteristics of lung lesions including size, density, and location make the detection of lung nodules on chest X-rays more challenging. However, machine learning methods, including the implementation of deep convolutional neural networks (DCNN), have helped to improve detection.


New AI deep learning model allows earlier, more accurate ozone warnings

#artificialintelligence

That would improve health alerts for people at heightened risk of developing problems because of high ozone levels. Yunsoo Choi, associate professor in the Department of Earth and Atmospheric Sciences and corresponding author for a paper explaining the work, said they built an artificially intelligent model using a convolutional neural network, which is able to take information from current conditions and accurately predict ozone levels for the next day. The work was published in the journal Neural Networks. "If we know the conditions of today, we can predict the conditions of tomorrow," Choi said. Ozone is an unstable gas, formed by a chemical reaction when sunlight combines with nitrogen oxides (NOx) and volatile organic compounds, both of which are found in automobile and industrial emissions.


Additive Powers-of-Two Quantization: A Non-uniform Discretization for Neural Networks

arXiv.org Machine Learning

We proposed Additive Powers-of-Two~(APoT) quantization, an efficient non-uniform quantization scheme that attends to the bell-shaped and long-tailed distribution of weights in neural networks. By constraining all quantization levels as a sum of several Powers-of-Two terms, APoT quantization enjoys overwhelming efficiency of computation and a good match with weights' distribution. A simple reparameterization on clipping function is applied to generate better-defined gradient for updating of optimal clipping threshold. Moreover, weight normalization is presented to refine the input distribution of weights to be more stable and consistent. Experimental results show that our proposed method outperforms state-of-the-art methods, and is even competitive with the full-precision models demonstrating the effectiveness of our proposed APoT quantization. For example, our 3-bit quantized ResNet-34 on ImageNet only drops 0.3% Top-1 and 0.2% Top-5 accuracy without bells and whistles, while the computation of our model is approximately 2x less than uniformly quantized neural networks.


Representation Learning for Electronic Health Records

arXiv.org Machine Learning

Information in electronic health records (EHR), such as clinical narratives, examination reports, lab measurements, demographics, and other patient encounter entries, can be transformed into appropriate data representations that can be used for downstream clinical machine learning tasks using representation learning. Learning better representations is critical to improve the performance of downstream tasks. Due to the advances in machine learning, we now can learn better and meaningful representations from EHR through disentangling the underlying factors inside data and distilling large amounts of information and knowledge from heterogeneous EHR sources. In this chapter, we first introduce the background of learning representations and reasons why we need good EHR representations in machine learning for medicine and healthcare in Section 1. Next, we explain the commonly-used machine learning and evaluation methods for representation learning using a deep learning approach in Section 2. Following that, we review recent related studies of learning patient state representation from EHR for clinical machine learning tasks in Section 3. Finally, in Section 4 we discuss more techniques, studies, and challenges for learning natural language representations when free texts, such as clinical notes, examination reports, or biomedical literature are used. W e also discuss challenges and opportunities in these rapidly growing research fields.


Introducing quantum convolutional neural networks

#artificialintelligence

Machine learning techniques have so far proved to be very promising for the analysis of data in several fields, with many potential applications. However, researchers have found that applying these methods to quantum physics problems is far more challenging due to the exponential complexity of many-body systems. Quantum many-body systems are essentially microscopic structures made up of several interacting particles. While quantum physics studies have focused on the collective behavior of these systems, using machine learning in these investigations has proven to be very difficult. With this in mind, a team of researchers at Harvard University recently developed a quantum circuit-based algorithm inspired by convolutional neural networks (CNNs), a popular machine learning technique that has achieved remarkable results in a variety of fields.


Introducing quantum convolutional neural networks

#artificialintelligence

Machine learning techniques have so far proved to be very promising for the analysis of data in several fields, with many potential applications. However, researchers have found that applying these methods to quantum physics problems is far more challenging due to the exponential complexity of many-body systems. Quantum many-body systems are essentially microscopic structures made up of several interacting particles. While quantum physics studies have focused on the collective behavior of these systems, using machine learning in these investigations has proven to be very difficult. With this in mind, a team of researchers at Harvard University recently developed a quantum circuit-based algorithm inspired by convolutional neural networks (CNNs), a popular machine learning technique that has achieved remarkable results in a variety of fields.


AI: Psychosensory electronic skin technology for future AI development

#artificialintelligence

As a result, many scientists are continuously performing research to imitate tactile, olfactory, and palate senses and tactile sensing is expected to be the next mimetic technology for various reasons. Currently, most tactile sensor researches are focusing on physical mimetic technologies that measure the pressure used for a robot to grab an object, but psychosensory tactile research on how to mimic human tactile feeling such like soft, smooth or rough has a long way to go. As a result, Professor Jae Eun Jang's team developed a tactile sensor that can feel pain and temperature like human through a joint research with Professor Cheil Moon's team in the Department of Brain and Cognitive Science, Professor Ji-woong Choi's team in the Department of Information and Communication Engineering, and Professor Hongsoo Choi's team in the Department of Robotics Engineering. Its key strengths are that it has simplified the sensor structure and can measure pressure and temperature at the same time and can be applied on various tactile systems regardless of the measurement principle of the sensor. For this, the research team focused on zinc oxide nano-wire (ZnO Nano-wire) technology, which was applied as a self-power tactile sensor that does not need a battery thanks to its piezoelectric effect, which generates electrical signals by detecting pressure.


Psychosensory electronic skin technology for future AI and humanoid development

#artificialintelligence

Professor Jae Eun Jang's team in the Department of Information and Communication Engineering has developed electronic skin technology that can detect "prick" and "hot" pain sensations like humans. This research result has applications in the development of humanoid robots and prosthetic hands in the future. Scientists are continuously performing research to imitate tactile, olfactory and palate senses, and tactile sensing is expected to be the next mimetic technology for various applications. Currently, most tactile sensor research is focused on physical mimetic technologies that measure the pressure used for a robot to grab an object, but psychosensory tactile research on mimicking human tactile sensory responses like those caused by soft, smooth or rough surfaces has a long way to go. Professor Jae Eun Jang's team has developed a tactile sensor that can feel pain and temperature like humans through a joint project with Professor Cheil Moon's team in the Department of Brain and Cognitive Science, Professor Ji-woong Choi's team in the Department of Information and Communication Engineering, and Professor Hongsoo Choi's team in the Department of Robotics Engineering.


Psychosensory electronic skin technology for future AI and humanoid development

#artificialintelligence

The attempt to mimic human's five senses led to the development of innovative electronic devices such as camera and TV, which are inventions that dramatically changed human life. As a result, many scientists are continuously performing research to imitate tactile, olfactory, and palate senses and tactile sensing is expected to be the next mimetic technology for various reasons. Currently, most tactile sensor researches are focusing on physical mimetic technologies that measure the pressure used for a robot to grab an object, but psychosensory tactile research on how to mimic human tactile feeling such like soft, smooth or rough has a long way to go. As a result, Professor Jae Eun Jang's team developed a tactile sensor that can feel pain and temperature like human through a joint research with Professor Cheil Moon's team in the Department of Brain and Cognitive Science, Professor Ji-woong Choi's team in the Department of Information and Communication Engineering, and Professor Hongsoo Choi's team in the Department of Robotics Engineering. Its key strengths are that it has simplified the sensor structure and can measure pressure and temperature at the same time and can be applied on various tactile systems regardless of the measurement principle of the sensor.