This is our 101'st blog post here on Learning Machines and we have prepared something very special for you! Oftentimes the different concepts of data science, namely artificial intelligence (AI), machine learning (ML), and deep learning (DL) are confused… so we asked the most advanced AI in the world, OpenAI GPT-3, to write a guest post for us to provide some clarification on their definitions and how they are related. We are most delighted to present this very impressive (and only slightly redacted) essay to you – enjoy! Artificial intelligence (AI), machine learning (ML), and deep learning (DL) are related concepts that are often used interchangeably. They are also three distinct and different concepts.

Historically, both economists and philosophers have been preoccupied with extracting an understanding of cause and effect from empirical evidence. David Hume, an economist and philosopher, is renowned for exploring causality, both as an epistemological puzzle and as a matter of practical concern in applied economics. In an article titled "Causality in Economics and Econometrics", economics professor Kevin D. Hoover states, "economists inherited from Hume the sense that practical economics was essentially a causal science." (Hoover, 2006). As a capital "E" Empiricist, Hume was a major influence on the development of causality in economics; his skepticism created a tension between the "epistemological status of causal relations and their role in practical policy." (Hoover, 2006). In 1739, when Hume famously defined causation in the "Treatise of Human Nature", I doubt he would have been able to foresee the radical change wrought by the exponential progress of technological evolution. Nor could he have imagined our present-day reality where deep learning is used to determine cause and effect. Today, "causal science" is being driven by machine learning, however, it is still a nascent area and development has focused primarily on theory.

Reservoir computing (RC) offers efficient temporal data processing with a low training cost by separating recurrent neural networks into a fixed network with recurrent connections and a trainable linear network. The quality of the fixed network, called reservoir, is the most important factor that determines the performance of the RC system. In this paper, we investigate the influence of the hierarchical reservoir structure on the properties of the reservoir and the performance of the RC system. Analogous to deep neural networks, stacking sub-reservoirs in series is an efficient way to enhance the nonlinearity of data transformation to high-dimensional space and expand the diversity of temporal information captured by the reservoir. These deep reservoir systems offer better performance when compared to simply increasing the size of the reservoir or the number of sub-reservoirs. Low frequency components are mainly captured by the sub-reservoirs in later stage of the deep reservoir structure, similar to observations that more abstract information can be extracted by layers in the late stage of deep neural networks. When the total size of the reservoir is fixed, tradeoff between the number of sub-reservoirs and the size of each sub-reservoir needs to be carefully considered, due to the degraded ability of individual sub-reservoirs at small sizes.

Many businesses and industries nowadays rely on large quantities of time series data making time series forecasting an important research area. Global forecasting models that are trained across sets of time series have shown a huge potential in providing accurate forecasts compared with the traditional univariate forecasting models that work on isolated series. However, there are currently no comprehensive time series archives for forecasting that contain datasets of time series from similar sources available for the research community to evaluate the performance of new global forecasting algorithms over a wide variety of datasets. In this paper, we present such a comprehensive time series forecasting archive containing 20 publicly available time series datasets from varied domains, with different characteristics in terms of frequency, series lengths, and inclusion of missing values. We also characterise the datasets, and identify similarities and differences among them, by conducting a feature analysis. Furthermore, we present the performance of a set of standard baseline forecasting methods over all datasets across eight error metrics, for the benefit of researchers using the archive to benchmark their forecasting algorithms.

Bayesian neural networks have shown great promise in many applications where calibrated uncertainty estimates are crucial and can often also lead to a higher predictive performance. However, it remains challenging to choose a good prior distribution over their weights. While isotropic Gaussian priors are often chosen in practice due to their simplicity, they do not reflect our true prior beliefs well and can lead to suboptimal performance. Our new library, BNNpriors, enables state-of-the-art Markov Chain Monte Carlo inference on Bayesian neural networks with a wide range of predefined priors, including heavy-tailed ones, hierarchical ones, and mixture priors. Moreover, it follows a modular approach that eases the design and implementation of new custom priors. It has facilitated foundational discoveries on the nature of the cold posterior effect in Bayesian neural networks and will hopefully catalyze future research as well as practical applications in this area.

We investigate molecular mechanisms of resistant or sensitive response of cancer drug combination therapies in an inductive and interpretable manner. Though deep learning algorithms are widely used in the drug synergy prediction problem, it is still an open problem to formulate the prediction model with biological meaning to investigate the mysterious mechanisms of synergy (MoS) for the human-AI collaboration in healthcare systems. To address the challenges, we propose a deep graph neural network, IDSP (Interpretable Deep Signaling Pathways), to incorporate the gene-gene as well as gene-drug regulatory relationships in synergic drug combination predictions. IDSP automatically learns weights of edges based on the gene and drug node relations, i.e., signaling interactions, by a multi-layer perceptron (MLP) and aggregates information in an inductive manner. The proposed architecture generates interpretable drug synergy prediction by detecting important signaling interactions, and can be implemented when the underlying molecular mechanism encounters unseen genes or signaling pathways. We test IDWSP on signaling networks formulated by genes from 46 core cancer signaling pathways and drug combinations from NCI ALMANAC drug combination screening data. The experimental results demonstrated that 1) IDSP can learn from the underlying molecular mechanism to make prediction without additional drug chemical information while achieving highly comparable performance with current state-of-art methods; 2) IDSP show superior generality and flexibility to implement the synergy prediction task on both transductive tasks and inductive tasks. 3) IDSP can generate interpretable results by detecting different salient signaling patterns (i.e. MoS) for different cell lines.

This paper is based on a machine learning project at the Norwegian University of Science and Technology, fall 2020. The project was initiated with a literature review on the latest developments within time-series forecasting methods in the scientific community over the past five years. The paper summarizes the essential aspects of this research. Furthermore, in this paper, we introduce an LSTM cell's architecture, and explain how different components go together to alter the cell's memory and predict the output. Also, the paper provides the necessary formulas and foundations to calculate a forward iteration through an LSTM. Then, the paper refers to some practical applications and research that emphasize the strength and weaknesses of LSTMs, shown within the time-series domain and the natural language processing (NLP) domain. Finally, alternative statistical methods for time series predictions are highlighted, where the paper outline ARIMA and exponential smoothing. Nevertheless, as LSTMs can be viewed as a complex architecture, the paper assumes that the reader has some knowledge of essential machine learning aspects, such as the multi-layer perceptron, activation functions, overfitting, backpropagation, bias, over- and underfitting, and more.

This paper is a part of a student project in Machine Learning at the Norwegian University of Science and Technology. In this paper, a deep convolutional neural network with five convolutional layers and three fully-connected layers is presented to estimate the ages of individuals based on images. The model is in its entirety trained from scratch, where a combination of three different datasets is used as training data. These datasets are the APPA dataset, UTK dataset, and the IMDB dataset. The images were preprocessed using a proprietary face-recognition software. Our model is evaluated on both a held-out test set, and on the Adience benchmark. On the test set, our model achieves a categorical accuracy of 52%. On the Adience benchmark, our model proves inferior compared with other leading models, with an exact accuray of 30%, and an one-off accuracy of 46%. Furthermore, a script was created, allowing users to estimate their age directly using their web camera. The script, alongside all other code, is located in our GitHub repository: AgeNet.

Synthetic data generation is of great interest in diverse applications, such as for privacy protection. Deep generative models, such as variational autoencoders (VAEs), are a popular approach for creating such synthetic datasets from original data. Despite the success of VAEs, there are limitations when it comes to the bimodal and skewed marginal distributions. These deviate from the unimodal symmetric distributions that are encouraged by the normality assumption typically used for the latent representations in VAEs. While there are extensions that assume other distributions for the latent space, this does not generally increase flexibility for data with many different distributions. Therefore, we propose a novel method, pre-transformation variational autoencoders (PTVAEs), to specifically address bimodal and skewed data, by employing pre-transformations at the level of original variables. Two types of transformations are used to bring the data close to a normal distribution by a separate parameter optimization for each variable in a dataset. We compare the performance of our method with other state-of-the-art methods for synthetic data generation. In addition to the visual comparison, we use a utility measurement for a quantitative evaluation. The results show that the PTVAE approach can outperform others in both bimodal and skewed data generation. Furthermore, the simplicity of the approach makes it usable in combination with other extensions of VAE.

In this paper, we study the problem of extracting variable-depth "logical document hierarchy" from long documents, namely organizing the recognized "physical document objects" into hierarchical structures. The discovery of logical document hierarchy is the vital step to support many downstream applications. However, long documents, containing hundreds or even thousands of pages and variable-depth hierarchy, challenge the existing methods. To address these challenges, we develop a framework, namely Hierarchy Extraction from Long Document (HELD), where we "sequentially" insert each physical object at the proper on of the current tree. Determining whether each possible position is proper or not can be formulated as a binary classification problem. To further improve its effectiveness and efficiency, we study the design variants in HELD, including traversal orders of the insertion positions, heading extraction explicitly or implicitly, tolerance to insertion errors in predecessor steps, and so on. The empirical experiments based on thousands of long documents from Chinese, English financial market and English scientific publication show that the HELD model with the "root-to-leaf" traversal order and explicit heading extraction is the best choice to achieve the tradeoff between effectiveness and efficiency with the accuracy of 0.9726, 0.7291 and 0.9578 in Chinese financial, English financial and arXiv datasets, respectively. Finally, we show that logical document hierarchy can be employed to significantly improve the performance of the downstream passage retrieval task. In summary, we conduct a systematic study on this task in terms of methods, evaluations, and applications.