Most of the results above were actually first published in TU Munich's FKI Tech Report series, for which I drew many illustrations by hand, some of them shown in the present page (Sec. The FKI series now plays an important role in the history of Artificial Intelligence, as it introduced several important concepts: Unsupervised Pre-Training for Very Deep Learning (FKI-148-91 [UN0], Sec. In particular, the report FKI-126-90 [AC90] introduced a whole bunch of concepts that are now widely used: Planning with Recurrent World Models (Sec. Later remarkable FKI Tech Reports from the 1990s describe ways of greatly compressing NNs [KO0] [FM] to improve their generalisation capability.

The interpretability of neural networks is becoming increasingly necessary, as deep learning is being adopted in settings where accurate and justifiable predictions are required. These applications range from finance to medical imaging. However, deep neural networks are notorious for a lack of justification. Explainable AI (XAI) attempts to bridge this divide between accuracy and interpretability, but as we explain below, XAI justifies decisions without interpreting the model directly. Defining explainability or interpretability for computer vision is challenging: What does it even mean to explain a classification for high-dimensional inputs like images? As we discuss below, two popular definitions involve saliency maps and decision trees, but both approaches have their weaknesses.

Summary: Explaining data science to a non-data scientist isn't as easy as it sounds. You may know a lot about math, tools, techniques, data, and computer architecture but the question is how do you explain this briefly without getting buried in the detail. You might try this approach. You're at a party or maybe striking up a conversation with that pretty girl at the bar and sooner or later the question comes up, "what do you do?" Since you have what is reported to be the sexiest job in the world you proudly respond "I'm a data scientist". OK, what happens next depends on exactly what you say.

Created by Lazy Programmer Team, Lazy Programmer Inc. English [Auto-generated] Created by Lazy Programmer Team, Lazy Programmer Inc. Welcome to PyTorch: Deep Learning and Artificial Intelligence! Although Google's Deep Learning library Tensorflow has gained massive popularity over the past few years, PyTorch has been the library of choice for professionals and researchers around the globe for deep learning and artificial intelligence. Is it possible that Tensorflow is popular only because Google is popular and used effective marketing? Why did Tensorflow change so significantly between version 1 and version 2? Was there something deeply flawed with it, and are there still potential problems? It is less well-known that PyTorch is backed by another Internet giant, Facebook (specifically, the Facebook AI Research Lab - FAIR).

The landscape of traditional retail is experiencing a seismic shift. A rapidly evolving competitive environment, a global move towards digital shopping, and the ever-changing sentiments of highly informed buyers are forcing a new perspective in the industry. This competitive situation is forcing traditional retailers to innovate and adopt accelerated analytics, robotics, and deep learning. At Drishtic we are focused on improving the intelligence at each retail store by leveraging the data from Video cameras already deployed in the stores.

Models based on neural networks and machine learning are seeing a rise in popularity in space physics. In particular, the forecasting of geomagnetic indices with neural network models is becoming a popular field of study. These models are evaluated with metrics such as the root-mean-square error (RMSE) and Pearson correlation coefficient. However, these classical metrics sometimes fail to capture crucial behavior. To show where the classical metrics are lacking, we trained a neural network, using a long short-term memory network, to make a forecast of the disturbance storm time index at origin time $t$ with a forecasting horizon of 1 up to 6 hours, trained on OMNIWeb data. Inspection of the model's results with the correlation coefficient and RMSE indicated a performance comparable to the latest publications. However, visual inspection showed that the predictions made by the neural network were behaving similarly to the persistence model. In this work, a new method is proposed to measure whether two time series are shifted in time with respect to each other, such as the persistence model output versus the observation. The new measure, based on Dynamical Time Warping, is capable of identifying results made by the persistence model and shows promising results in confirming the visual observations of the neural network's output. Finally, different methodologies for training the neural network are explored in order to remove the persistence behavior from the results.

Graph Attention Network (GAT) and GraphSAGE are neural network architectures that operate on graph-structured data and have been widely studied for link prediction and node classification. One challenge raised by GraphSAGE is how to smartly combine neighbour features based on graph structure. GAT handles this problem through attention, however the challenge with GAT is its scalability over large and dense graphs. In this work, we proposed a new architecture to address these issues that is more efficient and is capable of incorporating different edge type information. It generates node representations by attending to neighbours sampled from weighted multi-step transition probabilities. We conduct experiments on both transductive and inductive settings. Experiments achieved comparable or better results on several graph benchmarks, including the Cora, Citeseer, Pubmed, PPI, Twitter, and YouTube datasets.

Deep neural networks can now perform many tasks that were once thought to be only feasible for humans. Unfortunately, while reaching impressive performance under standard settings, such networks are known to be susceptible to adversarial attacks -- slight but carefully constructed perturbations of the inputs which drastically decrease the network performance and reduce their trustworthiness. Here we propose to improve network robustness to input perturbations via an adversarial training procedure which we call Adversarial Feature Desensitization (AFD). We augment the normal supervised training with an adversarial game between the embedding network and an additional adversarial decoder which is trained to discriminate between the clean and perturbed inputs from their high-level embeddings. Our theoretical and empirical evidence acknowledges the effectiveness of this approach in learning robust features on MNIST, CIFAR10, and CIFAR100 datasets -- substantially improving the state-of-the-art in robust classification against previously observed adversarial attacks. More importantly, we demonstrate that AFD has better generalization ability than previous methods, as the learned features maintain their robustness against a large range of perturbations, including perturbations not seen during training. These results indicate that reducing feature sensitivity using adversarial training is a promising approach for ameliorating the problem of adversarial attacks in deep neural networks.

Forecasting stock prices can be interpreted as a time series prediction problem, for which Long Short Term Memory (LSTM) neural networks are often used due to their architecture specifically built to solve such problems. In this paper, we consider the design of a trading strategy that performs portfolio optimization using the LSTM stock price prediction for four different companies. We then customize the loss function used to train the LSTM to increase the profit earned. Moreover, we propose a data driven approach for optimal selection of window length and multi-step prediction length, and consider the addition of analyst calls as technical indicators to a multi-stack Bidirectional LSTM strengthened by the addition of Attention units. We find the LSTM model with the customized loss function to have an improved performance in the training bot over a regressive baseline such as ARIMA, while the addition of analyst call does improve the performance for certain datasets.

We present a neural network architecture that is fully equivariant with respect to transformations under the Lorentz group, a fundamental symmetry of space and time in physics. The architecture is based on the theory of the finite-dimensional representations of the Lorentz group and the equivariant nonlinearity involves the tensor product. For classification tasks in particle physics, we demonstrate that such an equivariant architecture leads to drastically simpler models that have relatively few learnable parameters and are much more physically interpretable than leading approaches that use CNNs and point cloud approaches. The competitive performance of the network is demonstrated on a public classification dataset [27] for tagging top quark decays given energy-momenta of jet constituents produced in proton-proton collisions.