Split learning (splitNN) has emerged as a popular strategy for addressing the high computational costs and low modeling efficiency in Vertical Federated Learning (VFL). However, despite its popularity, vanilla splitNN lacks encryption protection, leaving it vulnerable to privacy leakage issues, especially Label Leakage from Gradients (LLG). Motivated by the LLG issue resulting from the use of labels during training, we propose the Generative Adversarial Federated Model (GAFM), a novel method designed specifically to enhance label privacy protection by integrating splitNN with Generative Adversarial Networks (GANs). GAFM leverages GANs to indirectly utilize label information by learning the label distribution rather than relying on explicit labels, thereby mitigating LLG. GAFM also employs an additional cross-entropy loss based on the noisy labels to further improve the prediction accuracy. Our ablation experiment demonstrates that the combination of GAN and the cross-entropy loss component is necessary to enable GAFM to mitigate LLG without significantly compromising the model utility. Empirical results on various datasets show that GAFM achieves a better and more robust trade-off between model utility and privacy compared to all baselines across multiple random runs. In addition, we provide experimental justification to substantiate GAFM's superiority over splitNN, demonstrating that it offers enhanced label protection through gradient perturbation relative to splitNN.
Structural magnetic resonance imaging (sMRI) is widely used for brain neurological disease diagnosis; while longitudinal MRIs are often collected to monitor and capture disease progression, as clinically used in diagnosing Alzheimer's disease (AD). However, most current methods neglect AD's progressive nature and only take a single sMRI for recognizing AD. In this paper, we consider the problem of leveraging the longitudinal MRIs of a subject for AD identification. To capture longitudinal changes in sMRIs, we propose a novel model Longformer, a spatiotemporal transformer network that performs attention mechanisms spatially on sMRIs at each time point and integrates brain region features over time to obtain longitudinal embeddings for classification. Our Longformer achieves state-of-the-art performance on two binary classification tasks of separating different stages of AD using the ADNI dataset. Our source code is available at https://github.com/Qybc/LongFormer.
ADHD is a prevalent disorder among the younger population. Standard evaluation techniques currently use evaluation forms, interviews with the patient, and more. However, its symptoms are similar to those of many other disorders like depression, conduct disorder, and oppositional defiant disorder, and these current diagnosis techniques are not very effective. Thus, a sophisticated computing model holds the potential to provide a promising diagnosis solution to this problem. This work attempts to explore methods to diagnose ADHD using combinations of multiple established machine learning techniques like neural networks and SVM models on the ADHD200 dataset and explore the field of neuroscience. In this work, multiclass classification is performed on phenotypic data using an SVM model. The better results have been analyzed on the phenotypic data compared to other supervised learning techniques like Logistic regression, KNN, AdaBoost, etc. In addition, neural networks have been implemented on functional connectivity from the MRI data of a sample of 40 subjects provided to achieve high accuracy without prior knowledge of neuroscience. It is combined with the phenotypic classifier using the ensemble technique to get a binary classifier. It is further trained and tested on 400 out of 824 subjects from the ADHD200 data set and achieved an accuracy of 92.5% for binary classification The training and testing accuracy has been achieved upto 99% using ensemble classifier.
Falls have become more frequent in recent years, which has been harmful for senior citizens.Therefore detecting falls have become important and several data sets and machine learning model have been introduced related to fall detection. In this project report, a human fall detection method is proposed using a multi modality approach. We used the UP-FALL detection data set which is collected by dozens of volunteers using different sensors and two cameras. We use wrist sensor with acclerometer data keeping labels to binary classification, namely fall and no fall from the data set.We used fusion of camera and sensor data to increase performance. The experimental results shows that using only wrist data as compared to multi sensor for binary classification did not impact the model prediction performance for fall detection.
We establish estimations for the parameters of the output distribution for the softmax activation function using the probit function. As an application, we develop a new efficient Bayesian learning algorithm for fully connected neural networks, where training and predictions are performed within the Bayesian inference framework in closed-form. This approach allows sequential learning and requires no computationally expensive gradient calculation and Monte Carlo sampling. Our work generalizes the Bayesian algorithm for a single perceptron for binary classification in \cite{H} to multi-layer perceptrons for multi-class classification.
Self-distillation (SD) is the process of first training a \enquote{teacher} model and then using its predictions to train a \enquote{student} model with the \textit{same} architecture. Specifically, the student's objective function is $\big(\xi*\ell(\text{teacher's predictions}, \text{ student's predictions}) + (1-\xi)*\ell(\text{given labels}, \text{ student's predictions})\big)$, where $\ell$ is some loss function and $\xi$ is some parameter $\in [0,1]$. Empirically, SD has been observed to provide performance gains in several settings. In this paper, we theoretically characterize the effect of SD in two supervised learning problems with \textit{noisy labels}. We first analyze SD for regularized linear regression and show that in the high label noise regime, the optimal value of $\xi$ that minimizes the expected error in estimating the ground truth parameter is surprisingly greater than 1. Empirically, we show that $\xi > 1$ works better than $\xi \leq 1$ even with the cross-entropy loss for several classification datasets when 50\% or 30\% of the labels are corrupted. Further, we quantify when optimal SD is better than optimal regularization. Next, we analyze SD in the case of logistic regression for binary classification with random label corruption and quantify the range of label corruption in which the student outperforms the teacher in terms of accuracy. To our knowledge, this is the first result of its kind for the cross-entropy loss.
Football is a very result-driven industry, with goals being rarer than in most sports, so having further parameters to judge the performance of teams and individuals is key. Expected Goals (xG) allow further insight than just a scoreline. To tackle the need for further analysis in football, this paper uses machine learning applications that are developed and applied to Football Event data. From the concept, a Binary Classification problem is created whereby a probabilistic valuation is outputted using Logistic Regression and Gradient Boosting based approaches. The model successfully predicts xGs probability values for football players based on 15,575 shots. The proposed solution utilises StatsBomb as the data provider and an industry benchmark to tune the models in the right direction. The proposed ML solution for xG is further used to tackle the age-old cliche of: 'the ball has fallen to the wrong guy there'. The development of the model is used to adjust and gain more realistic values of expected goals than the general models show. To achieve this, this paper tackles Positional Adjusted xG, splitting the training data into Forward, Midfield, and Defence with the aim of providing insight into player qualities based on their positional sub-group. Positional Adjusted xG successfully predicts and proves that more attacking players are better at accumulating xG. The highest value belonged to Forwards followed by Midfielders and Defenders. Finally, this study has further developments into Player Adjusted xG with the aim of proving that Messi is statistically at a higher efficiency level than the average footballer. This is achieved by using Messi subset samples to quantify his qualities in comparison to the average xG models finding that Messi xG performs 347 xG higher than the general model outcome.
Sparsity finds applications in areas as diverse as statistics, machine learning, and signal processing. Computations over sparse structures are less complex compared to their dense counterparts, and their storage consumes less space. This paper proposes a heuristic method for retrieving sparse approximate solutions of optimization problems via minimizing the $\ell_{p}$ quasi-norm, where $0<p<1$. An iterative two-block ADMM algorithm for minimizing the $\ell_{p}$ quasi-norm subject to convex constraints is proposed. For $p=s/q<1$, $s,q \in \mathbb{Z}_{+}$, the proposed algorithm requires solving for the roots of a scalar degree $2q$ polynomial as opposed to applying a soft thresholding operator in the case of $\ell_{1}$. The merit of that algorithm relies on its ability to solve the $\ell_{p}$ quasi-norm minimization subject to any convex set of constraints. However, it suffers from low speed, due to a convex projection step in each iteration, and the lack of mathematical convergence guarantee. We then aim to vanquish these shortcomings by relaxing the assumption on the constraints set to be the set formed due to convex and differentiable, with Lipschitz continuous gradient, functions, i.e. specifically, polytope sets. Using a proximal gradient step, we mitigate the convex projection step and hence enhance the algorithm speed while proving its convergence. We then present various applications where the proposed algorithm excels, namely, matrix rank minimization, sparse signal reconstruction from noisy measurements, sparse binary classification, and system identification. The results demonstrate the significant gains obtained by the proposed algorithm compared to those via $\ell_{1}$ minimization.
Accurate detection of brain tumors could save lots of lives and increasing the accuracy of this binary classification even as much as a few percent has high importance. Neural Gas Networks (NGN) is a fast, unsupervised algorithm that could be used in data clustering, image pattern recognition, and image segmentation. In this research, we used the metaheuristic Firefly Algorithm (FA) for image contrast enhancement as pre-processing and NGN weights for feature extraction and segmentation of Magnetic Resonance Imaging (MRI) data on two brain tumor datasets from the Kaggle platform. Also, tumor classification is conducted by Support Vector Machine (SVM) classification algorithms and compared with a deep learning technique plus other features in train and test phases. Additionally, NGN tumor segmentation is evaluated by famous performance metrics such as Accuracy, F-measure, Jaccard, and more versus ground truth data and compared with traditional segmentation techniques. The proposed method is fast and precise in both tasks of tumor classification and segmentation compared with other methods. A classification accuracy of 95.14 % and segmentation accuracy of 0.977 is achieved by the proposed method.
The Internet of Things (IoT) faces tremendous security challenges. Machine learning models can be used to tackle the growing number of cyber-attack variations targeting IoT systems, but the increasing threat posed by adversarial attacks restates the need for reliable defense strategies. This work describes the types of constraints required for a realistic adversarial cyber-attack example and proposes a methodology for a trustworthy adversarial robustness analysis with a realistic adversarial evasion attack vector. The proposed methodology was used to evaluate three supervised algorithms, Random Forest (RF), Extreme Gradient Boosting (XGB), and Light Gradient Boosting Machine (LGBM), and one unsupervised algorithm, Isolation Forest (IFOR). Constrained adversarial examples were generated with the Adaptative Perturbation Pattern Method (A2PM), and evasion attacks were performed against models created with regular and adversarial training. Even though RF was the least affected in binary classification, XGB consistently achieved the highest accuracy in multi-class classification. The obtained results evidence the inherent susceptibility of tree-based algorithms and ensembles to adversarial evasion attacks and demonstrates the benefits of adversarial training and a security by design approach for a more robust IoT network intrusion detection and cyber-attack classification.
Variance in predictions across different trained models is a significant, under-explored source of error in fair binary classification. In practice, the variance on some data examples is so large that decisions can be effectively arbitrary. To investigate this problem, we take an experimental approach and make four overarching contributions: We: 1) Define a metric called self-consistency, derived from variance, which we use as a proxy for measuring and reducing arbitrariness; 2) Develop an ensembling algorithm that abstains from classification when a prediction would be arbitrary; 3) Conduct the largest to-date empirical study of the role of variance (vis-a-vis self-consistency and arbitrariness) in fair binary classification; and, 4) Release a toolkit that makes the US Home Mortgage Disclosure Act (HMDA) datasets easily usable for future research. Altogether, our experiments reveal shocking insights about the reliability of conclusions on benchmark datasets. Most fair binary classification benchmarks are close-to-fair when taking into account the amount of arbitrariness present in predictions -- before we even try to apply any fairness interventions. This finding calls into question the practical utility of common algorithmic fairness methods, and in turn suggests that we should reconsider how we choose to measure fairness in binary classification.
With the expansion of AI-powered virtual assistants, there is a need for low-power keyword spotting systems providing a "wake-up" mechanism for subsequent computationally expensive speech recognition. One promising approach is the use of neuromorphic sensors and spiking neural networks (SNNs) implemented in neuromorphic processors for sparse event-driven sensing. However, this requires resource-efficient SNN mechanisms for temporal encoding, which need to consider that these systems process information in a streaming manner, with physical time being an intrinsic property of their operation. In this work, two candidate neurocomputational elements for temporal encoding and feature extraction in SNNs described in recent literature - the spiking time-difference encoder (TDE) and disynaptic excitatory-inhibitory (E-I) elements - are comparatively investigated in a keyword-spotting task on formants computed from spoken digits in the TIDIGITS dataset. While both encoders improve performance over direct classification of the formant features in the training data, enabling a complete binary classification with a logistic regression model, they show no clear improvements on the test set. Resource-efficient keyword spotting applications may benefit from the use of these encoders, but further work on methods for learning the time constants and weights is required to investigate their full potential.
We look at the internal structure of neural networks which is usually treated as a black box. The easiest and the most comprehensible thing to do is to look at a binary classification and try to understand the approach a neural network takes. We review the significance of different activation functions, types of network architectures associated to them, and some empirical data. We find some interesting observations and a possibility to build upon the ideas to verify the process for real datasets. We suggest some possible experiments to look forward to in three different directions.
Large-scale object detection and instance segmentation face a severe data imbalance. The finer-grained object classes become, the less frequent they appear in our datasets. However, at test-time, we expect a detector that performs well for all classes and not just the most frequent ones. In this paper, we provide a theoretical understanding of the long-trail detection problem. We show how the commonly used mean average precision evaluation metric on an unknown test set is bound by a margin-based binary classification error on a long-tailed object detection training set. We optimize margin-based binary classification error with a novel surrogate objective called \textbf{Effective Class-Margin Loss} (ECM). The ECM loss is simple, theoretically well-motivated, and outperforms other heuristic counterparts on LVIS v1 benchmark over a wide range of architecture and detectors. Code is available at \url{https://github.com/janghyuncho/ECM-Loss}.
Deterministic solutions are becoming more critical for interpretability. Weighted Least-Squares (WLS) has been widely used as a deterministic batch solution with a specific weight design. In the online settings of WLS, exact reweighting is necessary to converge to its batch settings. In order to comply with its necessity, the iteratively reweighted least-squares algorithm is mainly utilized with a linearly growing time complexity which is not attractive for online learning. Due to the high and growing computational costs, an efficient online formulation of reweighted least-squares is desired. We introduce a new deterministic online classification algorithm of WLS with a constant time complexity for binary class rebalancing. We demonstrate that our proposed online formulation exactly converges to its batch formulation and outperforms existing state-of-the-art stochastic online binary classification algorithms in real-world data sets empirically.
Since last 2 decades, High Frequency Oscillations (HFOs) are studied as a promising biomarker to localize the epileptogenic zone of patients with refractory focal epilepsy. As HFOs visual detection is time consuming and subjective, automatization of HFO detection is required. Most HFO detectors were developed on invasive electroencephalograms (iEEG) whereas scalp electroencephalograms (EEG) are used in clinical routine. In order HFO detection can benefit to more patients, scalp HFO detectors has to be developed. However, HFOs identification in scalp EEG is more challenging than in iEEG since scalp HFOs are of lower rate, lower amplitude and more likely to be corrupted by several sources of artifacts than iEEG HFOs. The main goal of this study is to explore the ability of deep learning architecture to identify scalp HFOs from the remaining EEG signal. Hence, a binary classification Convolutional Neural Network (CNN) is learned to analyze High Density Electroencephalograms (HD-EEG). EEG signals are first mapped into a 2D time-frequency image, several color definitions are then used as an input for the CNN. Experimental results show that deep learning allows simple end-to-end learning of preprocessing, feature extraction and classification modules while reaching competitive performance.
This paper explores connections between margin-based loss functions and consistency in binary classification and regression applications. It is shown that a large class of margin-based loss functions for binary classification/regression result in estimating scores equivalent to log-likelihood scores weighted by an even function. A simple characterization for conformable (consistent) loss functions is given, which allows for straightforward comparison of different losses, including exponential loss, logistic loss, and others. The characterization is used to construct a new Huber-type loss function for the logistic model. A simple relation between the margin and standardized logistic regression residuals is derived, demonstrating that all margin-based loss can be viewed as loss functions of squared standardized logistic regression residuals. The relation provides new, straightforward interpretations for exponential and logistic loss, and aids in understanding why exponential loss is sensitive to outliers. In particular, it is shown that minimizing empirical exponential loss is equivalent to minimizing the sum of squared standardized logistic regression residuals. The relation also provides new insight into the AdaBoost algorithm.
We study the performance of machine learning binary classification techniques in terms of error probabilities. The statistical test is based on the Data-Driven Decision Function (D3F), learned in the training phase, i.e., what is thresholded before the final binary decision is made. Based on large deviations theory, we show that under appropriate conditions the classification error probabilities vanish exponentially, as $\sim \exp\left(-n\,I + o(n) \right)$, where $I$ is the error rate and $n$ is the number of observations available for testing. We also propose two different approximations for the error probability curves, one based on a refined asymptotic formula (often referred to as exact asymptotics), and another one based on the central limit theorem. The theoretical findings are finally tested using the popular MNIST dataset.
Human Activity Recognition (HAR) enables context-aware user experiences where mobile apps can alter content and interactions depending on user activities. Hence, smartphones have become valuable for HAR as they allow large, and diversified data collection. Although previous work in HAR managed to detect simple activities (i.e., sitting, walking, running) with good accuracy using inertial sensors (i.e., accelerometer), the recognition of complex daily activities remains an open problem, specially in remote work/study settings when people are more sedentary. Moreover, understanding the everyday activities of a person can support the creation of applications that aim to support their well-being. This paper investigates the recognition of complex activities exclusively using smartphone accelerometer data. We used a large smartphone sensing dataset collected from over 600 users in five countries during the pandemic and showed that deep learning-based, binary classification of eight complex activities (sleeping, eating, watching videos, online communication, attending a lecture, sports, shopping, studying) can be achieved with AUROC scores up to 0.76 with partially personalized models. This shows encouraging signs toward assessing complex activities only using phone accelerometer data in the post-pandemic world.
Speech patterns have been identified as potential diagnostic markers for neuropsychiatric conditions. However, most studies only compare a single clinical group to healthy controls, whereas clinical practice often requires differentiating between multiple potential diagnoses (multiclass settings). To address this, we assembled a dataset of repeated recordings from 420 participants (67 with major depressive disorder, 106 with schizophrenia and 46 with autism, as well as matched controls), and tested the performance of a range of conventional machine learning models and advanced Transformer models on both binary and multiclass classification, based on voice and text features. While binary models performed comparably to previous research (F1 scores between 0.54-0.75 for autism spectrum disorder, ASD; 0.67-0.92 for major depressive disorder, MDD; and 0.71-0.83 for schizophrenia); when differentiating between multiple diagnostic groups performance decreased markedly (F1 scores between 0.35-0.44 for ASD, 0.57-0.75 for MDD, 0.15-0.66 for schizophrenia, and 0.38-0.52 macro F1). Combining voice and text-based models yielded increased performance, suggesting that they capture complementary diagnostic information. Our results indicate that models trained on binary classification may learn to rely on markers of generic differences between clinical and non-clinical populations, or markers of clinical features that overlap across conditions, rather than identifying markers specific to individual conditions. We provide recommendations for future research in the field, suggesting increased focus on developing larger transdiagnostic datasets that include more fine-grained clinical features, and that can support the development of models that better capture the complexity of neuropsychiatric conditions and naturalistic diagnostic assessment.