Machine Learning Today Research Articles

Identifying, Mitigating, and Anticipating Bias in Algorithmic Decisions

Today's machine learning (ML) applications predominantly adhere to a standard paradigm: the decision maker designs the algorithm by optimizing a model for some objective function. While this has proven to be a powerful approach in many domains, it comes with inherent side effects: the power over the algorithmic outcomes lies solely in the hands of the algorithm designer, and alternative objectives, such as fairness, are often disregarded. This is particularly problematic if the algorithm is used to make consequential decisions that affect peoples lives. My research focuses on developing principled methods to characterize and address the mismatch between these different objectives.

From Statistical Relational to Neuro-Symbolic Artificial Intelligence

The integration of learning and reasoning is one of the key challenges in artificial intelligence and machine learning today. The area of Neuro-Symbolic AI (NeSy) tackles this challenge by integrating symbolic reasoning with neural networks. In our recent work, we provided an introduction to NeSy by drawing several parallels to another field that has a rich tradition in integrating learning and reasoning, namely Statistical Relational Artificial Intelligence (StarAI).

Computational biology is poised to advance precision medicine with machine learning

Computational biology is poised to advance precision medicine with machine learning Today, scientists are attempting to model whole cells using computational biology, building virtual cells that capture the dynamics of living. In the post-genomic era, vast quantities of data describing the parts of living cells, both normally functioning and diseased, have been amassed. Understanding how cells behave has been the goal of reductionist science, and many of the principles are well understood. In elucidating the details, complexity has been revealed to the degree that scientists have limited capacity to understand the systems underlying the biological networks. This is to say, cellular behavior is difficult to predict.

Two Applications of Manifold Regularization in Deep Learning Architectures

With the rapid development of machine learning today, deep learning has shown great advantages in many fields. Due to the high cost of datasets with labels, its branch semi-supervised learning has become an important research area. However, existing semi-supervised learning methods perform poorly on extremely small datasets, for instance, only one labeled example for each class. Hence, we introduce methods that apply manifold approaches to DL(Deep Learning) and frameworks that implement the idea. Potential application scenarios were analyzed based on experiments on synthetic and real-world experiments.

Embryo Transfer is the last Frontier for Deep Machine Learning & Artificial Intelligence in Medically Assisted Reproduction (MAR)

Embryo Transfer is the last Frontier for Deep Machine Learning & Artificial Intelligence in Medically Assisted Reproduction (MAR). In the last five years, nearly every aspect of an IVF cycle has been investigated by artificial intelligence(AI) & deep machine learning (ML), including sperm morphology, sperm identification, identification of empty or oocyte containing follicles, prediction of embryo cell stages, prediction of blastulation from oocytes, scoring blastocyst quality, prediction of euploid blastocysts and live birth from blastocysts, improving the embryo selection process, and for developing algorithms for optimal IVF stimulation protocols. Moreover, AI-based methods can be implemented for other clinical aspects of IVF, such as assessing patient reproductive potential and individualizing gonadotropin stimulation protocols. As AI has the capability to analyze "big" data, the ultimate goal will be to apply AI tools to the analysis of all embryological, clinical, and genetic data in an effort to provide patient-tailored treatments. Embryo Transfer is the only step of IVF that is outside the realm of AI & ML today. Embryo Transfer success is presently human skill dependent and deep machine learning may one day intrude into this sacred space with the advent of specialized humanoid robots. Embryo transfer is arguably the rate limiting step in the sequential events that complete an IVF cycle. Many variables play a role in the success of embryo transfer, including catheter type, atraumatic technique, and the use of sonography guidance. In this clinical review we will cover the contemporary research goals of AI & ML as well as the variables influencing Embryo Transfer success.

Positive climate effects when AR customer support simultaneous trains AI experts for the smart industries of the future

Initially, Artificial Intelligence (AI) focused on diagnostics during the 70s and 80s. Unfortunately, it did not gain trust and few industries embraced it, mostly due to the extensive manual programming effort that AI required for interpreting data and act. In addition, the computer capacity, for handling the amounts of data necessary to train AI, was lacking the disc dimensions we are used to today, which made it go slowly. Not until the 2000 s confidence in AI was established in parallel with the introduction of new tools that was paving the way for PLS, PCA, ANN and soft sensors. Year 2011, IBM Watson (an AI application) was developed and won over the jeopardy champion. Today's machine learning (ML) such as “deep learning” and artificial neural networks (ANN) have created interesting use cases. AI has therefore regained confidence and industries are beginning to embrace where they see appropriate uses. Simultaneously, Internet of Things (IoT) tools have been introduced and made it possible to develop new capabilities such as virtual reality (VR), augmented reality (AR), mixed reality (MR) and extended reality (XR). These technologies are maturing and could be used in several application areas for the industries and form part of their digitalization journey. Furthermore, it is not only the industries that could benefit from introducing these technologies. Studies also show several areas and use cases where augmented reality has a positive impact, such as on students' learning ability. Yet few teachers know or use this technology. This paper evaluates and analyze AR, remote assistance tool for industrial purposes. The potential of the tool is discussed for frequent maintenance cases in the mining industry. Further on, if we look into the future, it is not surprising if we will be able to see that today's concepts of reality tools have evolved to become smarter by being trained by multimedia recognition and from people who have thus created an AI expert. Where the AI expert will support customers and be able to solve simple errors but also those that occur rarely and thus be a natural part of the solution for future completely autonomous processes for the industry. The article demonstrates a framework for creating smarter tools by combining AR, ML and AI and forms part of the basis creating the smarter industry of the future. Natural Language Processing (NLP) toolbox has been utilized to train and test an AI expert to give suitable resolutions to a specific maintenance request. The motivation for AR is the possible energy savings and reduction of CO₂ emissions in the maintenance field for all business trips that can be avoided. At the same time saving money for the industries and expert manhours that are spent on traveling and finally enhancing the productivity for the industries. Tests cases have verified that with AR, the resolution time could be significantly reduced, minimizing production stoppages by more than 50% of the time, which ultimately has a positive effect on a country's GDP. How much energy can be saved is predicted by the fact that 50% of all the world's business flights are replaced by one of the reality concepts and are estimated to amount to at least 50 Mton CO₂ per year. This figure is probably slightly higher as business trips also take place by other means of transport such as trains, buses, and cars. With today's volatile employees changing jobs more frequently, industry experts are becoming fewer and fewer. Since new employee stays for a maximum of 3–5 years per workplace, they will not stay long enough to become experts. Introducing an AI expert trained by today's experts, there is a chance that this knowledge can be maintained.

MACHINE LEARNING ALGORITHMS FOR DETECTION OF CYBER THREATS USING LOGISTIC REGRESSION

The threat of cyberattacks is expanding globally; thus, businesses are developing intelligent artificial intelligence systems that can analyze security and other infrastructure logs from their systems department and quickly and automatically identify cyberattacks. Security analytics based on machine learning the next big thing in cybersecurity is machine data, which aims to mine security data to show the high maintenance costs of static relationship rules and methods. But, choosing the appropriate machine learning technique for log analytics using ML continues to be a significant barrier to AI success in cyber security due to the possibility of a substantial number of false-positive detections in large-scale or global Security Operations Centre (SOC) settings, selecting the proper machine learning technique for security log analytics remains a substantial obstacle to AI success in cyber security. A machine learning technique for a cyber threat exposure system that can minimize false positives is required. Today's machine learning methods for identifying threats frequently use logistic regression. Logistic regression is the first of three machine learning subcategories—supervised, unsupervised, and reinforcement learning. Any machine learning enthusiast will encounter this supervised machine learning algorithm at the beginning of their machine learning career. It's an essential and often applied classification algorithm.

Is Quantum Advantage the Right Goal for Quantum Machine Learning?

Machine learning is frequently listed among the most promising applications for quantum computing. This is in fact a curious choice: Today's machine learning algorithms are notoriously powerful in practice, but remain theoretically difficult to study. Quantum computing, in contrast, does not offer practical benchmarks on realistic scales, and theory is the main tool we have to judge whether it could become relevant for a problem. In this perspective we explain why it is so difficult to say something about the practical power of quantum computers for machine learning with the tools we are currently using. We argue that these challenges call for a critical debate on whether quantum advantage and the narrative of 'beating' classical machine learning should continue to dominate the literature the way it does, and highlight examples for how other perspectives in existing research provide an important alternative to the focus on advantage.

Survey: Tensorflow in Machine Learning

TensorFlow(TF) is a large-scale ML system that can be used in many different situations. Data-flow graphs are used by TF to define operations, shared state, and computations that change that state. It distributes the nodes of a Data-flow graph across multiple machines in a cluster, as well as diverse processing units within a single system, such as multicore CPUs, general-purpose GPUs and Tensor Processing Units, which are custom-designed ASICs (TPUs). Developers may use TF to try out new training techniques and optimizations. It covers a broad applications range, which focus on inference and DNN training. It is utilised in production by several Google services, and it has since been released as an open-source project and is broadly used in ML research. We review TensorFlow and place it in the context of modern deep learning techniques and technologies in this paper. We go over the fundamental computational concepts and how they work. The distributed execution concept, as well as its programming interface as accompanying visualisation toolkit. The remainder of this paper will provide a comprehensive overview of TensorFlow and place it in the context of machine learning today

GrapHD: Graph-Based Hyperdimensional Memorization for Brain-Like Cognitive Learning.

Memorization is an essential functionality that enables today's machine learning algorithms to provide a high quality of learning and reasoning for each prediction. Memorization gives algorithms prior knowledge to keep the context and define confidence for their decision. Unfortunately, the existing deep learning algorithms have a weak and nontransparent notion of memorization. Brain-inspired HyperDimensional Computing (HDC) is introduced as a model of human memory. Therefore, it mimics several important functionalities of the brain memory by operating with a vector that is computationally tractable and mathematically rigorous in describing human cognition. In this manuscript, we introduce a brain-inspired system that represents HDC memorization capability over a graph of relations. We propose GrapHD, hyperdimensional memorization that represents graph-based information in high-dimensional space. GrapHD defines an encoding method representing complex graph structure while supporting both weighted and unweighted graphs. Our encoder spreads the information of all nodes and edges across into a full holistic representation so that no component is more responsible for storing any piece of information than another. Then, GrapHD defines several important cognitive functionalities over the encoded memory graph. These operations include memory reconstruction, information retrieval, graph matching, and shortest path. Our extensive evaluation shows that GrapHD: (1) significantly enhances learning capability by giving the notion of short/long term memorization to learning algorithms, (2) enables cognitive computing and reasoning over memorization graph, and (3) enables holographic brain-like computation with substantial robustness to noise and failure.

Machine Learning on Volatile Instances: Convergence, Runtime, and Cost Tradeoffs

Due to the massive size of the neural network models and training datasets used in machine learning today, it is imperative to distribute stochastic gradient descent (SGD) by splitting up tasks such as gradient evaluation across multiple worker nodes. However, running distributed SGD can be prohibitively expensive because it may require specialized computing resources such as GPUs for extended periods of time. We propose cost-effective strategies to exploit volatile cloud instances that are cheaper than standard instances, but may be interrupted by higher priority workloads. To the best of our knowledge, this work is the first to quantify how variations in the number of active worker nodes (as a result of preemption) affect SGD convergence and the time to train the model. By understanding these trade-offs between preemption probability of the instances, accuracy, and training time, we are able to derive practical strategies for configuring distributed SGD jobs on volatile instances such as Amazon EC2 spot instances and other preemptible cloud instances. Experimental results show that our strategies achieve good training performance at substantially lower cost.

Systematic Literature Review of Waste Classification Using Machine Learning

The development of the global economy has caused people's living standards to increase and the production of domestic waste has also increased from year to year. The population of big cities that have limited environmental carrying capacity, causing the waste problem requires serious handling. Manual waste sorting is hazardous to health and wastes time, money and effort. If waste is not handled properly, environmental problems will increase in the long run. Machine learning works by combining features such as textures and colors to complement junk image recognition. Today's machine learning technology continues to develop, not only methods, types of waste, and features but also identify and analyze datasets used in waste management by gathering all scientific evidence. Collecting existing research and then identifying, assessing, and interpreting requires a systematic literature review. Until the end of 2021, the research topic of waste classification using machine learning was found with various types of waste, algorithms, datasets, and others. However, the dataset used by the algorithm in image recognition is relatively single, the types of garbage classified and the relative accuracy results can still be improved.

Robotic platform equipped with machine learning

Automatic lines equipped with stationary robots are a key element of the industry. The robots are integrated into production lines, to meet basic, repetitive operations, with a finite degree of variability in internal programs. Reprogramming in terms of, for example, changing a manufactured, manipulated part is time-consuming and cost-effective. However, the development of today's machine learning algorithms is only carefully integrated in this market segment. Manufacturers do not provide their closed systems with a sufficient degree of programming variability. The solution tries to outline this work, which complements the standard industrial robot with a cognitive interface. Such a robot is able to learn new programs and make production changes on the fly.

PROBLEMS OF RETRO-FORECASTS OF ECONOMIC INDICATORS BASED ON NEURAL NETWORKS

Introduction. TThe difficult economic situation in Ukraine, the state of emergency, the war and the consequences of the destruction of critical infrastructure have significantly increased the value of economic and mathematical forecasting tools based on neural networks and their tuning capabilities, is improved. The purpose of the article is to analyze and study the potential and formal aspects of the application of neurocomputer methods of economic forecasting and tools to support retro-forecasts of economic indicators. Results. The success of economic decisions (strategic and tactical) in one way or another depends on the quality of analytics and the efficiency of the operational apparatus of decision-making. Even if a rigorous algorithmic approach is difficult or impossible and it is fundamentally impossible to get the right solution, there are effective methods and tools for solving economic problems, an important place among which is forecasting. Just in recent years there has been a breakthrough in the theory and practice of economics and mathematics, many organizations have begun to actively use neural networks in forecasting. Such neural networks can identify patterns by which they generate recommendations for action, they can study and summarize past experiences to improve their own level of performance and calculate forecasts. Neural network methodologies in general and neuro-forecasting in particular belong to the family of machine learning technologies. The specifics of solving the problem of machine learning in forecasting differs from other methods designed for the formation and use of predictions. Possibilities of accumulation and improvement of experience, formation and adaptation of neural network architecture to specifics of forecasting tasks, wide possibilities of application of retro-forecasting methods for improvement of characteristics of adjustment of neural models for forecasts, availability of flexible mechanisms of parameterization and optimization of algorithmic providing forecasting in economic research. Machine learning today is a field of scientific knowledge that is developing rapidly and deals with algorithms capable of learning and developing, which distinguishes this field of research from many others and makes it, at the same time focused on implementation practice, including forecasting. The need to use machine learning methods is due to the fact that for many complex – “intelligent” – tasks (eg, construction and improvement of predictive models) is very difficult (or even impossible) to develop an “explicit” algorithm for solving them, but often you can teach a computer to learn solving these problems with the help of neural networks and retro-forecasting technologies. Conclusions. The formalization of the integrated representation of the neural network for forecasting in the form of activation functions with the definition of conditions for their use in machine learning algorithms, taking into account the specifics of the refined settings of retro-predicted neural models of economic indicators. Demonstrated by the example of retro-forecasting for the simplest neural network (4-2-3-1), when the choice of machine learning algorithm (between PROP and RPROP) is achieved more than twice the effect of improving the quality of the forecast model.

Design and Assessment of Convolutional Neural Network Based Methods for Vitiligo Diagnosis.

Background: Today's machine-learning based dermatologic research has largely focused on pigmented/non-pigmented lesions concerning skin cancers. However, studies on machine-learning-aided diagnosis of depigmented non-melanocytic lesions, which are more difficult to diagnose by unaided eye, are very few.Objective: We aim to assess the performance of deep learning methods for diagnosing vitiligo by deploying Convolutional Neural Networks (CNNs) and comparing their diagnosis accuracy with that of human raters with different levels of experience.Methods: A Chinese in-house dataset (2,876 images) and a world-wide public dataset (1,341 images) containing vitiligo and other depigmented/hypopigmented lesions were constructed. Three CNN models were trained on close-up images in both datasets. The results by the CNNs were compared with those by 14 human raters from four groups: expert raters (>10 years of experience), intermediate raters (5–10 years), dermatology residents, and general practitioners. F1 score, the area under the receiver operating characteristic curve (AUC), specificity, and sensitivity metrics were used to compare the performance of the CNNs with that of the raters.Results: For the in-house dataset, CNNs achieved a comparable F1 score (mean [standard deviation]) with expert raters (0.8864 [0.005] vs. 0.8933 [0.044]) and outperformed intermediate raters (0.7603 [0.029]), dermatology residents (0.6161 [0.068]) and general practitioners (0.4964 [0.139]). For the public dataset, CNNs achieved a higher F1 score (0.9684 [0.005]) compared to the diagnosis of expert raters (0.9221 [0.031]).Conclusion: Properly designed and trained CNNs are able to diagnose vitiligo without the aid of Wood's lamp images and outperform human raters in an experimental setting.

Living systems are smarter bots: Slime mold semiosis versus AI symbol manipulation

Although machines may be good at mimicking, they are not currently able, as organisms are, to act creatively. We offer an understanding of the emergent qualities of biological sign processing in terms of generalization, association, and encryption. We use slime mold as a model of minimal cognition and compare it to deep-learning video game bots, which some claim have evolved beyond their merely quantitative algorithms. We find that these discrete Turing machine bots are not able to make productive, yet unanticipated, “errors”—necessary for biological learning—which, based on the physicality of signs, their relatively similar shapes, and relative physical positions spatially and temporally, lead to emergent effects and make learning and evolution possible. In organisms, stochastic resonance at the local level can be leveraged for self-organization at the global level. We contrast all this to the symbolic processing of today's machine learning, whereby each logic node and memory state is discrete. Computer codes are produced by external operators, whereas biological symbols are evolved through an internal encryption process.

Higgs analysis with quantum classifiers

We have developed two quantum classifier models for the ttH classification problem, both of which fall into the category of hybrid quantumclassical algorithms for Noisy Intermediate Scale Quantum devices (NISQ). Our results, along with other studies, serve as a proof of concept that Quantum Machine Learning (QML) methods can have similar or better performance, in specific cases of low number of training samples, with respect to conventional ML methods even with a limited number of qubits available in current hardware. To utilise algorithms with a low number of qubits — to accommodate for limitations in both simulation hardware and real quantum hardware — we investigated different feature reduction methods. Their impact on the performance of both the classical and quantum models was assessed. We addressed different implementations of two QML models, representative of the two main approaches to supervised quantum machine learning today: a Quantum Support Vector Machine (QSVM), a kernel-based method, and a Variational Quantum Circuit (VQC), a variational approach.

Quantum Algorithms for Feedforward Neural Networks

Quantum machine learning has the potential for broad industrial applications, and the development of quantum algorithms for improving the performance of neural networks is of particular interest given the central role they play in machine learning today. We present quantum algorithms for training and evaluating feedforward neural networks based on the canonical classical feedforward and backpropagation algorithms. Our algorithms rely on an efficient quantum subroutine for approximating inner products between vectors in a robust way, and on implicitly storing intermediate values in quantum random access memory for fast retrieval at later stages. The running times of our algorithms can be quadratically faster in the size of the network than their standard classical counterparts since they depend linearly on the number of neurons in the network, and not on the number of connections between neurons. Furthermore, networks trained by our quantum algorithm may have an intrinsic resilience to overfitting, as the algorithm naturally mimics the effects of classical techniques used to regularize networks. Our algorithms can also be used as the basis for new quantum-inspired classical algorithms with the same dependence on the network dimensions as their quantum counterparts but with quadratic overhead in other parameters that makes them relatively impractical.

A Primer on Machine Learning.

A Primer on Machine Learning.

Privacy-Preserving Machine Learning: Threats and Solutions

For privacy concerns to be addressed adequately in today's machine-learning (ML) systems, the knowledge gap between the ML and privacy communities must be bridged. This article aims to provide an introduction to the intersection of both fields with special emphasis on the techniques used to protect the data.