Artificial Intelligence is the latest buzzword in the technology space. As expected, the automation has made things easier and smarter in various industries, ranging from healthcare, startups to even human resources. Unlike the implementation of Artificial Intelligence in other sectors, the HR space becomes trickier as it involves sensitive employee-employer relationship. Also, will employees trust machines to evaluate their performance or even play a role in recruiting them? Nonetheless, there are several advantages that AI brings to this segment.

Karthik Padmanabhan is the Developer Relations Program lead at Google and is responsible for India, Middle East and North Africa regions. Karthik has been with the tech industry for almost three decades in the areas of product management, business development, tech evangelism, etc. He leads a team that focuses on enabling developer communities to adopt Google & open source technologies like TensorFlow, PWA, Android, & Google Cloud for the Next Billion Users (NBU). Karthik is a seeker, plays Golf, works at Google's Bangalore office and lives on a farm that functions on sustainable living practices. For more details visit http://devfest.gdgahmedabad.com/

Artificial intelligence is a tool humanity is wielding with increasing recklessness. We say it's for our common good with machine learning hype equal to business profits. But what happens when we don't have the code of ethics, laws, government accountability, corporate transparency and capability of monitoring the space to be able to achieve AI regulation? Artificial intelligence regulation isn't just complex terrain, it's uncharted territory for an age that is passing the baton from human leadership to machine learning emergence, automation, robotic manufacturing and deep learning reliance. Artificial intelligence (AI) is an area of computer science that emphasizes the creation of intelligent machines that work and react like humans.

Are you consciously or unconsciously aware of the fact that Artificial Intelligence is omnipresent? It is finely weaved in our day to day routine from phones to computer and tablets, every device embraces the technology in our surrounding. Even the rising trending craze for Netflix is also a gift of Artificial Intelligence, for sure. Therefore, it would not be an element surprise if we monitor the involvement of the Indian government in this sector. The Government of India is set to regulate Artificial Intelligence with a transparent set of guidelines for the procedure to develop and implement the AI technology, as confirmed by Minister of Commerce & Industry and Civil Aviation Suresh Prabhu.

Parent company Alphabet's Google and Verily are using machine knowledge to help screen diabetic retinopathy (DR) and diabetic macular edema (DME). The two eye conditions are one of the major causes of blindness and if everything works well, it may facilitate automated screening. Which means that it will be able to detect disease sooner and provide more people with access to screenings. A part of Google's parent company Alphabet called Verily is working with Google to conduct clinical research around the world, especially in India, where studies showed the algorithm was as good at assessing images for disease as general ophthalmologists and retinal specialists, according to a blog post from the companies. Artificial Intelligence Can Detect Heart Problems in the Future, Says Study by Mayo Clinic. The "First real-world clinical use of the algorithm" has been initiated at Aravind Eye Hospital in Madurai as a result of the research.

When the weights in a particle filter are not available analytically, standard resampling methods cannot be employed. To circumvent this problem state-of-the-art algorithms replace the true weights with non-negative unbiased estimates. This algorithm is still valid but at the cost of higher variance of the resulting filtering estimates in comparison to a particle filter using the true weights. We propose here a novel algorithm that allows for resampling according to the true intractable weights when only an unbiased estimator of the weights is available. We demonstrate our algorithm on several examples.

Reinforcement Learning agents are expected to eventually perform well. Typically, this takes the form of a guarantee about the asymptotic behavior of an algorithm given some assumptions about the environment. We present an algorithm for a policy whose value approaches the optimal value with probability 1 in all computable probabilistic environments, provided the agent has a bounded horizon. This is known as strong asymptotic optimality, and it was previously unknown whether it was possible for a policy to be strongly asymptotically optimal in the class of all computable probabilistic environments. Our agent, Inquisitive Reinforcement Learner (Inq), is more likely to explore the more it expects an exploratory action to reduce its uncertainty about which environment it is in, hence the term inquisitive. Exploring inquisitively is a strategy that can be applied generally; for more manageable environment classes, inquisitiveness is tractable. We conducted experiments in "grid-worlds" to compare the Inquisitive Reinforcement Learner to other weakly asymptotically optimal agents.

Recently, deep generative models have become increasingly popular in unsupervised anomaly detection. However, deep generative models aim at recovering the data distribution rather than detecting anomalies. Besides, deep generative models have the risk of overfitting training samples, which has disastrous effects on anomaly detection performance. To solve the above two problems, we propose a Self-adversarial Variational Autoencoder with a Gaussian anomaly prior assumption. We assume that both the anomalous and the normal prior distribution are Gaussian and have overlaps in the latent space. Therefore, a Gaussian transformer net T is trained to synthesize anomalous but near-normal latent variables. Keeping the original training objective of Variational Autoencoder, besides, the generator G tries to distinguish between the normal latent variables and the anomalous ones synthesized by T, and the encoder E is trained to discriminate whether the output of G is real. These new objectives we added not only give both G and E the ability to discriminate but also introduce additional regularization to prevent overfitting. Compared with the SOTA baselines, the proposed model achieves significant improvements in extensive experiments. Datasets and our model are available at a Github repository.

In likelihood-free settings where likelihood evaluations are intractable, approximate Bayesian computation (ABC) addresses the formidable inference task to discover plausible parameters of simulation programs that explain the observations. However, they demand large quantities of simulation calls. Critically, hyperparameters that determine measures of simulation discrepancy crucially balance inference accuracy and sample efficiency, yet are difficult to tune. In this paper, we present kernel embedding likelihood-free inference (KELFI), a holistic framework that automatically learns model hyperparameters to improve inference accuracy given limited simulation budget. By leveraging likelihood smoothness with conditional mean embeddings, we nonparametrically approximate likelihoods and posteriors as surrogate densities and sample from closed-form posterior mean embeddings, whose hyperparameters are learned under its approximate marginal likelihood. Our modular framework demonstrates improved accuracy and efficiency on challenging inference problems in ecology.

Spatio-temporal prediction plays an important role in many application areas especially in traffic domain. However, due to complicated spatio-temporal dependency and high non-linear dynamics in road networks, traffic prediction task is still challenging. Existing works either exhibit heavy training cost or fail to accurately capture the spatio-temporal patterns, also ignore the correlation between distant roads that share the similar patterns. In this paper, we propose a novel deep learning framework to overcome these issues: 3D Temporal Graph Convolutional Networks (3D-TGCN). Two novel components of our model are introduced. (1) Instead of constructing the road graph based on spatial information, we learn it by comparing the similarity between time series for each road, thus providing a spatial information free framework. (2) We propose an original 3D graph convolution model to model the spatio-temporal data more accurately. Empirical results show that 3D-TGCN could outperform state-of-the-art baselines.