Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userA Standardized Radiograph-Agnostic Framework and Platform For Evaluating AI Radiological Systems
58
0
0.0
(
0
)
Added by
Darlington Ahiale Akogo
Publication date
2020
fields
Informatics Engineering
and research's language is
English
Authors
Darlington Ahiale Akogo
Ask ChatGPT about the research
No Arabic abstract
Radiology has been essential to accurately diagnosing diseases and assessing responses to treatment. The challenge however lies in the shortage of radiologists globally. As a response to this, a number of Artificial Intelligence solutions are being developed. The challenge Artificial Intelligence radiological solutions however face is the lack of a benchmarking and evaluation standard, and the difficulties of collecting diverse data to truly assess the ability of such systems to generalise and properly handle edge cases. We are proposing a radiograph-agnostic platform and framework that would allow any Artificial Intelligence radiological solution to be assessed on its ability to generalise across diverse geographical location, gender and age groups.
rate research
Read More
With the growing complexity of deep learning methods adopted in practical applications, there is an increasing and stringent need to explain and interpret the decisions of such methods. In this work, we focus on explainable AI and propose a novel generic and model-agnostic framework for synthesizing input exemplars that maximize a desired response from a machine learning model. To this end, we use a generative model, which acts as a prior for generating data, and traverse its latent space using a novel evolutionary strategy with momentum updates. Our framework is generic because (i) it can employ any underlying generator, e.g. Variational Auto-Encoders (VAEs) or Generative Adversarial Networks (GANs), and (ii) it can be applied to any input data, e.g. images, text samples or tabular data. Since we use a zero-order optimization method, our framework is model-agnostic, in the sense that the machine learning model that we aim to explain is a black-box. We stress out that our novel framework does not require access or knowledge of the internal structure or the training data of the black-box model. We conduct experiments with two generative models, VAEs and GANs, and synthesize exemplars for various data formats, image, text and tabular, demonstrating that our framework is generic. We also employ our prototype synthetization framework on various black-box models, for which we only know the input and the output formats, showing that it is model-agnostic. Moreover, we compare our framework (available at https://github.com/antoniobarbalau/exemplar) with a model-dependent approach based on gradient descent, proving that our framework obtains equally-good exemplars in a shorter computational time.
In a world increasingly dominated by AI applications, an understudied aspect is the carbon and social footprint of these power-hungry algorithms that require copious computation and a trove of data for training and prediction. While profitable in the short-term, these practices are unsustainable and socially extractive from both a data-use and energy-use perspective. This work proposes an ESG-inspired framework combining socio-technical measures to build eco-socially responsible AI systems. The framework has four pillars: compute-efficient machine learning, federated learning, data sovereignty, and a LEEDesque certificate. Compute-efficient machine learning is the use of compressed network architectures that show marginal decreases in accuracy. Federated learning augments the first pillars impact through the use of techniques that distribute computational loads across idle capacity on devices. This is paired with the third pillar of data sovereignty to ensure the privacy of user data via techniques like use-based privacy and differential privacy. The final pillar ties all these factors together and certifies products and services in a standardized manner on their environmental and social impacts, allowing consumers to align their purchase with their values.
Disturbances in space weather can negatively affect several fields, including aviation and aerospace, satellites, oil and gas industries, and electrical systems, leading to economic and commercial losses. Solar flares are the most significant events that can affect the Earths atmosphere, thus leading researchers to drive efforts on their forecasting. The related literature is comprehensive and holds several systems proposed for flare forecasting. However, most techniques are tailor-made and designed for specific purposes, not allowing researchers to customize them in case of changes in data input or in the prediction algorithm. This paper proposes a framework to design, train, and evaluate flare prediction systems which present promising results. Our proposed framework involves model and feature selection, randomized hyper-parameters optimization, data resampling, and evaluation under operational settings. Compared to baseline predictions, our framework generated some proof-of-concept models with positive recalls between 0.70 and 0.75 for forecasting $geq M$ class flares up to 96 hours ahead while keeping the area under the ROC curve score at high levels.
Generative Adversarial Networks (GANs) have shown remarkable results in modeling complex distributions, but their evaluation remains an unsettled issue. Evaluations are essential for: (i) relative assessment of different models and (ii) monitoring the progress of a single model throughout training. The latter cannot be determined by simply inspecting the generator and discriminator loss curves as they behave non-intuitively. We leverage the notion of duality gap from game theory to propose a measure that addresses both (i) and (ii) at a low computational cost. Extensive experiments show the effectiveness of this measure to rank different GAN models and capture the typical GAN failure scenarios, including mode collapse and non-convergent behaviours. This evaluation metric also provides meaningful monitoring on the progression of the loss during training. It highly correlates with FID on natural image datasets, and with domain specific scores for text, sound and cosmology data where FID is not directly suitable. In particular, our proposed metric requires no labels or a pretrained classifier, making it domain agnostic.
In this paper, we report about a large-scale online discussion with 1099 citizens on the Afghanistan Sustainable Development Goals.
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
