When designing and building data pipelines to load data into data warehouses you might have heard of the common ETL and ELT paradigms. This post goes over what they mean, their differences and which paradigm you might want to choose. If you are wondering why we have a staging area click here. ELT is very similar but the data is loaded into a table before being transformed to a final table which is used by users. As you can see it has fewer components compared to the ETL approach.

Visual captioning aims to generate textual descriptions given images. Traditionally, the captioning models are trained on human annotated datasets such as Flickr30k and MS-COCO, which are limited in size and diversity. This limitation hinders the generalization capabilities of these models while also rendering them to often make mistakes. Language models can, however, be trained on vast amounts of freely available unlabelled data and have recently emerged as successful language encoders and coherent text generators. Meanwhile, several unimodal and multimodal fusion techniques have been proven to work well for natural language generation and automatic speech recognition. Building on these recent developments, and with an aim of improving the quality of generated captions, the contribution of our work in this paper is two-fold: First, we propose a generic multimodal model fusion framework for caption generation as well as emendation where we utilize different fusion strategies to integrate a pretrained Auxiliary Language Model (AuxLM) within the traditional encoder-decoder visual captioning frameworks. Next, we employ the same fusion strategies to integrate a pretrained Masked Language Model (MLM), namely BERT, with a visual captioning model, viz. Show, Attend, and Tell, for emending both syntactic and semantic errors in captions. Our caption emendation experiments on three benchmark image captioning datasets, viz. Flickr8k, Flickr30k, and MSCOCO, show improvements over the baseline, indicating the usefulness of our proposed multimodal fusion strategies. Further, we perform a preliminary qualitative analysis on the emended captions and identify error categories based on the type of corrections.

Researchers at the National University of Singapore recently demonstrated the advantages of using neuromorphic sensor fusion to help robots grip and identify objects. It's just one of a number of interesting projects they've been working on including developing a new protocol for transmitting tactile data, building a neuromorphic tactile fingertip, and developing new visual-tactile datasets for the development of better learning systems. Because the technology uses address-events and spiking neural networks it is extremely power efficient: 50 times more using one of the Intel Loihi neuromorphic chips than a GPU. However, what's particularly elegant about this work is that it points the way towards neuromorphic technology as a means of efficiently integrating -- and extracting meaning from -- many different sensors for complex tasks in power-constrained systems. The new tactile sensor they used, NeuTouch, consists of an array of 39 taxels (tactile pixels) and the movement is transduced using a graphene-based piezo-resistive layer; you can think as this as the front of the robot's fingertip.

The global artificial intelligence in energy market size is poised to grow by USD 8.06 billion during 2020-2024, decelerating at a CAGR of almost 48% …

At FutureLearn we work in short sprints & regularly share, reflect on and iterate on our work. This helps us focus on shipping small, iterative changes and responding quickly to changing business or user needs. We care about work/life balance and supporting learning at work. The Data Platform Team builds and maintains tooling and infrastructure that supports decision making processes across the business and enables product improvements by providing a complete and consistent view of our business data. Our tech stack consists of an ETL process written in Ruby and managed by Airflow which sources data from our production database (MySQL), our email provider (Sendgrid), application logs, and other operational data sources.

Learning the cumulative distribution function (CDF) of an outcome variable conditional on a set of features remains challenging, especially in high-dimensional settings. Conditional transformation models provide a semi-parametric approach that allows to model a large class of conditional CDFs without an explicit parametric distribution assumption and with only a few parameters. Existing estimation approaches within the class of transformation models are, however, either limited in their complexity and applicability to unstructured data sources such as images or text, or can incorporate complex effects of different features but lack interpretability. We close this gap by introducing the class of deep conditional transformation models which unify existing approaches and allow to learn both interpretable (non-)linear model terms and more complex predictors in one holistic neural network. To this end we propose a novel network architecture, provide details on different model definitions and derive suitable constraints and derive suitable network regularization terms. We demonstrate the efficacy of our approach through numerical experiments and applications.

This note is a little break from our model homotopy series. I have a neat example where one combines two classifiers to get a better classifier using a method I am calling "ROC surgery." In ROC surgery we look at multiple ROC plots and decide we want to cut out a section from one the plots for use. It is a sensor fusion method to try and combine the best parts of two classifiers.

Oracle is sprucing up its customer data platform, CX Unity, with a little machine learning. The platform will now support real-time behavioral data collection and personalization capabilities through Infinity, Oracle's digital streaming technology. Infinity captures web event, app and point-of-sale data to help brands build realistic representations of the customer journey. That previously required repeated and rigorous A/B testing, said Rob Tarkoff, EVP and general manager of Oracle Cloud CX and Oracle Data Cloud. "But by applying machine learning to that, you can come up with predictions and insights based on a holistic set of data, and it doesn't require you to do one-off activities," he said.

Data Science and Machine Learning can play a huge role in your organisation, our aim is to help you take off the load of the data discovery process. Our team would help you mine data, integrate it, and perform analytics with it and design suited data models that would aid your business.

Opioid overdose rates have increased in the United States over the past decade and reflect a major public health crisis. Modeling and prediction of drug and opioid hotspots, where a high percentage of events fall in a small percentage of space-time, could help better focus limited social and health services. In this work we present a spatial-temporal point process model for drug overdose clustering. The data input into the model comes from two heterogeneous sources: 1) high volume emergency medical calls for service (EMS) records containing location and time, but no information on the type of non-fatal overdose and 2) fatal overdose toxicology reports from the coroner containing location and high-dimensional information from the toxicology screen on the drugs present at the time of death. We first use non-negative matrix factorization to cluster toxicology reports into drug overdose categories and we then develop an EM algorithm for integrating the two heterogeneous data sets, where the mark corresponding to overdose category is inferred for the EMS data and the high volume EMS data is used to more accurately predict drug overdose death hotspots. We apply the algorithm to drug overdose data from Indianapolis, showing that the point process defined on the integrated data outperforms point processes that use only homogeneous EMS (AUC improvement .72 to .8) or coroner data (AUC improvement .81 to .85).We also investigate the extent to which overdoses are contagious, as a function of the type of overdose, while controlling for exogenous fluctuations in the background rate that might also contribute to clustering. We find that drug and opioid overdose deaths exhibit significant excitation, with branching ratio ranging from .72 to .98.