Everyone in the modern era, where the internet is widely used, relies on a range of online sources for news channels. Because more people are using Facebook, Instagram, and other social media platforms, news has sprea...
详细信息
Models produced by machinelearning are not guaranteed to be free from bias, particularly when trained and tested with data produced in discriminatory environments. The bias can be unethical, mainly when the data cont...
详细信息
ISBN:
(纸本)9781450394666
Models produced by machinelearning are not guaranteed to be free from bias, particularly when trained and tested with data produced in discriminatory environments. The bias can be unethical, mainly when the data contains sensitive attributes, such as sex, race, age, etc. Some approaches have contributed to mitigating such biases by providing bias metrics and mitigation algorithms. The challenge is users have to implement their code in general/statistical programming languages, which can be demanding for users with little programming and fairness in machinelearning experience. We present FairML, a model-based approach to facilitate bias measurement and mitigation with reduced software development effort. Our evaluation shows that FairML requires fewer lines of code to produce comparable measurement values to the ones produced by the baseline code.
Recent years, reinforcement learning has been developed in many area and achieve high performance in many application. However, for huge amount data and training will make hard leaning for the agent of reinforcement l...
详细信息
The goal of object detection is to recognize the position and category of all objects in an image, allowing for machine vision understanding. Many approaches have been developed to solve this problem, primarily based ...
详细信息
Due to its achievements in recent years, machinelearning (ML) is now used in a wide variety of domains. Educating ML has hence become an important factor in academia and industry. We argue that students learning abou...
详细信息
ISBN:
(纸本)9781665444347
Due to its achievements in recent years, machinelearning (ML) is now used in a wide variety of domains. Educating ML has hence become an important factor in academia and industry. We argue that students learning about machinelearning will need, in addition to theoretical knowledge, approaches to interactively explore machinelearning models and their parameters. This paper introduces EduML - an interactive approach for lecturers to teach and for students or professionals to study and explore the fundamentals of machinelearning. EduML allows users to experiment with data preparation, dimensionality reduction and a wide range of classifiers on different data sets. These data sets can be analysed in order to understand the complexity of the classification problem. The classifiers can be autonomously fitted to the training data or the effect of manually altering model hyperparameters can be explored. Additionally, to get started with programming own ML pipelines, Python and R source code of configured ML pipelines can be extracted. EduML has been used in a lecture as an interactive demo or by students in lab sessions. Roth scenarios were evaluated with a user survey.
machinelearning (ML) has been adopted in many safety-critical applications like automated driving and medical diagnosis. Incorrect decisions by ML models can lead to catastrophic consequences, such as vehicle crashes...
详细信息
ISBN:
(数字)9781665416931
ISBN:
(纸本)9781665416931
machinelearning (ML) has been adopted in many safety-critical applications like automated driving and medical diagnosis. Incorrect decisions by ML models can lead to catastrophic consequences, such as vehicle crashes and inappropriate medical procedures, thereby endangering our lives. The correct behaviour of a ML model is contingent upon the availability of well-labelled training data. However, obtaining large and high-quality training datasets for safety-critical applications is difficult, often resulting in the use of faulty training data. We compare the efficacy of five different error mitigation techniques, derived from a survey of more than 200 related articles, which are designed to tolerate noisy/faulty training data. We experimentally find that the error mitigation capabilities of these techniques vary across datasets, ML models, and different kinds of faults. We further find that ensemble learning offers the highest resilience among all the techniques across different configurations, followed by label smoothing.
Highly intelligent robot is the trend in the new round of development, and it is also the ultimate goal of robot navigation technology *** ability to learn is an important embodiment of robot *** navigation is the bas...
详细信息
Crisis management exercises play a key role in preparation for different scenarios such as natural disasters, or large crowd events. Moreover, they act as a platform for testing new technologies and finding common pra...
详细信息
Intrusion in an IoT (Internet of Things) device or an IoT based data is quite common but the data being shared in a secured manner is the point to be analysed. Any data that has a connection to the internet has a chan...
详细信息
Medical Named Entity Recognition (NER) aims to automatically identify entities like diseases, drugs, and symptoms from medical texts, supporting medical knowledge graphs, clinical decision-making, and intelligent syst...
详细信息
ISBN:
(纸本)9798400711848
Medical Named Entity Recognition (NER) aims to automatically identify entities like diseases, drugs, and symptoms from medical texts, supporting medical knowledge graphs, clinical decision-making, and intelligent systems. Current research relies on deep learning models, particularly pre-trained language models like BERT, improving recognition accuracy. Challenges include data sparsity, ambiguous boundaries, and synonym diversity, affecting generalization on specific datasets. To address this, we enhance BERT with prompt learning and contrastive learning, using learnable entity class embeddings and similarity computations to improve classification and recognition on few-shot datasets. Our model achieves F1 scores of 90.19% on the CCKS 2019 dataset and 83.30% on the JNLPBA dataset.
暂无评论