With the fire hose of imagery that's streaming daily from a variety of sensors, the need for using artificial intelligence (AI) to automate feature extraction is only increasing. The ability to train more than a dozen deep learning models on geospatial datasets and derive information products has been available using the ArcGIS API for Python or ArcGIS Pro, and users can scale up processing using ArcGIS Image Server. Esri is taking AI to the next level with ready-to-use geospatial AI models in the ArcGIS Living Atlas of the World. Initially, three models have been made available. Two of the models use satellite imagery.

Kafka's "The Metamorphosis" has a famous opening: "One morning, when Gregor Samsa woke from troubled dreams, he found himself transformed in his bed into a horrible vermin." The rest of the story follows, logically and ludicrously, from that original degrading miracle. Gregor struggles to get out of bed. His mother tells him that it's time to go to work. His boss, the chief clerk, shows up and demands that he return to the business no matter what shape he's in.

Being one of the leading high-level neural network APIs, Keras that is defined for human beings and not machines is written in python and supports multiple backends for neural network computation engines like Tensorflow, CNTK, Theano, MXNet, and PlaidML. A very powerful and easy-to-use free open source python library, Keras develops and evaluates deep learning models by wrapping around the functionalities of other ML and DL libraries like Theano and Tensorflow defining and training neural network models in just a few lines of code. Following the best practices of reducing cognitive load, Keras allows the users to productize deep models on smartphones, allows the usage of distributed training of deep learning models on clusters of GPU and TPU, and acts as an interface for the Tensorflow library making it easier to navigate.

We all know how quick was the Fourth Hokage. Known as Yellow Flash in all the five great nations whose rasengan made us all go "woooah". Well, we all can do the same with our deep learning models as well. This counters one of the biggest issue we face while doing deep learning projects, "LOW FPS". Passing frames from multiple videos or stream sources to our deep learning model causes slowdown and it is annoying as hell.

We first wrote this microsite in April 2017 specifically as a starting place for business and technical people just beginning their journeys with artificial intelligence. Since we first published this microsite in April 2017, there have been a number of breakthroughs and advances in the field of AI. In 2020, we saw the launch of GPT-3, the AlphaFold breakthrough in protein folding, the continued rise of data systems, and more. While we have not updated this guide to reflect all of these breakthrough, it remains a great initial guide for anyone involved in creating, managing, or green lighting software and want to understand how AI can give your software super powers (as Andrew Ng is fond of saying). We do not assume you have a computer science degree.

As mentioned above, this experiment uses the home insurance dataset. This dataset includes home insurance policy data between 2007 and 2012 and there are more than 100 features available regarding home features, owner's demographics, etc and there are more than 250,000 rows in this data. Using this dataset, this experiment tries to predict whether a home insurance policy is going to lapse. Unfortunately, not all the details of variables in this dataset were given, yet it is good enough to do this experiment.

Prediction based on Irregularly Sampled Time Series (ISTS) is of wide concern in the real-world applications. For more accurate prediction, the methods had better grasp more data characteristics. Different from ordinary time series, ISTS is characterised with irregular time intervals of intra-series and different sampling rates of inter-series. However, existing methods have suboptimal predictions due to artificially introducing new dependencies in a time series and biasedly learning relations among time series when modeling these two characteristics. In this work, we propose a novel Time Encoding (TE) mechanism. TE can embed the time information as time vectors in the complex domain. It has the the properties of absolute distance and relative distance under different sampling rates, which helps to represent both two irregularities of ISTS. Meanwhile, we create a new model structure named Time Encoding Echo State Network (TE-ESN). It is the first ESNs-based model that can process ISTS data. Besides, TE-ESN can incorporate long short-term memories and series fusion to grasp horizontal and vertical relations. Experiments on one chaos system and three real-world datasets show that TE-ESN performs better than all baselines and has better reservoir property.

The availability of massive earth observing satellite data provide huge opportunities for land use and land cover mapping. However, such mapping effort is challenging due to the existence of various land cover classes, noisy data, and the lack of proper labels. Also, each land cover class typically has its own unique temporal pattern and can be identified only during certain periods. In this article, we introduce a novel architecture that incorporates the UNet structure with Bidirectional LSTM and Attention mechanism to jointly exploit the spatial and temporal nature of satellite data and to better identify the unique temporal patterns of each land cover. We evaluate this method for mapping crops in multiple regions over the world. We compare our method with other state-of-the-art methods both quantitatively and qualitatively on two real-world datasets which involve multiple land cover classes. We also visualise the attention weights to study its effectiveness in mitigating noise and identifying discriminative time period.

To improve the physical understanding and the predictions of complex dynamic systems, such as ocean dynamics and weather predictions, it is of paramount interest to identify interpretable models from coarsely and off-grid sampled observations. In this work we investigate how deep learning can improve model discovery of partial differential equations when the spacing between sensors is large and the samples are not placed on a grid. We show how leveraging physics informed neural network interpolation and automatic differentiation, allow to better fit the data and its spatiotemporal derivatives, compared to more classic spline interpolation and numerical differentiation techniques. As a result, deep learning based model discovery allows to recover the underlying equations, even when sensors are placed further apart than the data's characteristic length scale and in the presence of high noise levels. We illustrate our claims on both synthetic and experimental data sets where combinations of physical processes such as (non)-linear advection, reaction and diffusion are correctly identified. Mathematical models are central in modelling complex dynamical processes such as climate change, the spread of an epidemic or to design aircrafts.

Knowledge Tracing (KT), tracking a human's knowledge acquisition, is a central component in online learning and AI in Education. In this paper, we present a simple, yet effective strategy to improve the generalization ability of KT models: we propose three types of novel data augmentation, coined replacement, insertion, and deletion, along with corresponding regularization losses that impose certain consistency or monotonicity biases on the model's predictions for the original and augmented sequence. Extensive experiments on various KT benchmarks show that our regularization scheme consistently improves the model performances, under 3 widely-used neural networks and 4 public benchmarks, e.g., it yields 6.3% improvement in AUC under the DKT model and the ASSISTmentsChall dataset.