In order apply the change-of-variables formula to get a density for the generator, we assume that G: Rd RD spans an immersedd-dimensional manifold inRD. The governing assumption is that the Jacobian ofG exist and has full rank. However, we note that one requirement is that no hidden layer may have dimensionality below the d dimensions of the latent space. This is a natural requirement for the generator anyway. In our model, we aim to maximize the entropy of the generator, which encourages the generator to create as diverse samples as possible.

We propose a new GAN-based unsupervised model for disentangled representation learning. The new model is discovered in an attempt to utilize the Information Bottleneck (IB) framework to the optimization of GAN, thereby named IB-GAN. The architecture of IB-GAN is partially similar to that of InfoGAN but has a critical difference; an intermediate layer of the generator is leveraged to constrain the mutual information between the input and the generated output. The intermediate stochastic layer can serve as a learnable latent distribution that is trained with the generator jointly in an end-to-end fashion. As a result, the generator of IB-GAN can harness the latent space in a disentangled and interpretable manner. With the experiments on dSprites and Color-dSprites dataset, we demonstrate that IB-GAN achieves competitive disentanglement scores to those of state-of-the-art \b{eta}-VAEs and outperforms InfoGAN. Moreover, the visual quality and the diversity of samples generated by IB-GAN are often better than those by \b{eta}-VAEs and Info-GAN in terms of FID score on CelebA and 3D Chairs dataset.

A.1 Proof of Theorem 1 Proof log null E The first inequality is derived by Holder's inequality, so Existence is ensured as long as the chosen activation functions have at least one derivative almost everywhere. Smooth activations naturally satisfy this assumption, but it is worth noting that e.g. the ReLU activation We cannot guarantee that the Jacobian has full rank through clever choices of neural architectures. This is a natural requirement for the generator anyway. In our model, we aim to maximize the entropy of the generator, which encourages the generator to create as diverse samples as possible. In practice this ensures that the Jacobian has full rank as a degenerate Jacobian implies a reduction of entropy.

Our examination of deep generative models (DGMs) developed for semi-supervised learning (SSL), mainly GANs and VAEs, reveals two problems. First, mode missing and mode covering phenomenons are observed in genertion with GANs and VAEs. Second, there exists an awkward conflict between good classification and good generation in SSL by employing directed generative models. To address these problems, we formally present joint-stochastic-approximation random fields (JRFs) -- a new family of algorithms for building deep undirected generative models, with application to SSL. It is found through synthetic experiments that JRFs work well in balancing mode covering and mode missing, and match the empirical data distribution well. Empirically, JRFs achieve good classification results comparable to the state-of-art methods on widely adopted datasets -- MNIST, SVHN, and CIFAR-10 in SSL, and simultaneously perform good generation.

In lossy image compression, the objective is to achieve minimal signal distortion while compressing images to a specified bit rate. The increasing demand for visual analysis applications, particularly in classification tasks, has emphasized the significance of considering semantic distortion in compressed images. To bridge the gap between image compression and visual analysis, we propose a Rate-Distortion-Classification (RDC) model for lossy image compression, offering a unified framework to optimize the trade-off between rate, distortion, and classification accuracy. The RDC model is extensively analyzed both statistically on a multi-distribution source and experimentally on the widely used MNIST dataset. The findings reveal that the RDC model exhibits desirable properties, including monotonic non-increasing and convex functions, under certain conditions. This work provides insights into the development of human-machine friendly compression methods and Video Coding for Machine (VCM) approaches, paving the way for end-to-end image compression techniques in real-world applications.

The field of text classification, a fundamental subdomain within the natural language processing (NLP) field of machine learning (ML), has witnessed a remarkable evolution in recent years. With the exponential increase in textual data generated across various domains, the need for effective text classification methods has become increasingly pressing. Text classification is the task of assigning predefined labels or categories to textual documents based on their content. This task holds immense importance across various industries and applications, including but not limited to sentiment analysis, spam detection, content recommendation, and news classification. The ability to automatically organize and categorize large volumes of text can streamline information retrieval, enhance decision-making processes, and enable efficient data management. Traditional text classification methods rely on well-established techniques such as term frequency - inverse document frequency (TF-IDF) representations and traditional ML algorithms. TF-IDF measures the importance of each term within a document relative to a corpus of documents, providing a numerical representation of textual data.

Autoencoders have been extensively used in the development of recent anomaly detection techniques. The premise of their application is based on the notion that after training the autoencoder on normal training data, anomalous inputs will exhibit a significant reconstruction error. Consequently, this enables a clear differentiation between normal and anomalous samples. In practice, however, it is observed that autoencoders can generalize beyond the normal class and achieve a small reconstruction error on some of the anomalous samples. To improve the performance, various techniques propose additional components and more sophisticated training procedures. In this work, we propose a remarkably straightforward alternative: instead of adding neural network components, involved computations, and cumbersome training, we complement the reconstruction loss with a computationally light term that regulates the norm of representations in the latent space. The simplicity of our approach minimizes the requirement for hyperparameter tuning and customization for new applications which, paired with its permissive data modality constraint, enhances the potential for successful adoption across a broad range of applications. We test the method on various visual and tabular benchmarks and demonstrate that the technique matches and frequently outperforms alternatives. We also provide a theoretical analysis and numerical simulations that help demonstrate the underlying process that unfolds during training and how it can help with anomaly detection. This mitigates the black-box nature of autoencoder-based anomaly detection algorithms and offers an avenue for further investigation of advantages, fail cases, and potential new directions.

We present TFF, which is a Transformer framework for the analysis of functional Magnetic Resonance Imaging (fMRI) data. TFF employs a two-phase training approach. First, self-supervised training is applied to a collection of fMRI scans, where the model is trained to reconstruct 3D volume data. Second, the pre-trained model is fine-tuned on specific tasks, utilizing ground truth labels. Our results show state-of-the-art performance on a variety of fMRI tasks, including age and gender prediction, as well as schizophrenia recognition. Our code for the training, network architecture, and results is attached as supplementary material.