The most important competency is the ability to learn quickly in a fast paced startup environment.

Yeah, I mean it "Awesome!". Deep learning is the most awesome, exciting, and interesting stuff I feel every data scientist should brag about. When I started my data science journey, I thought Learning basic machine learning was all until I heard about things like neural networks, NLP, Computer vision, and so forth. I was so eager to learn how deep learning works. However, getting started can be hectic and overwhelming since resources are all over the web, and getting the right resources can be difficult unless you have someone already practicing it to guide you, else you can waste a lot of time without getting started.

Depending on the initial conditions used, such as nutrient levels and space constraints, bacteria tend to grow in specific ways. The researchers created a virtual bacterial colony and then controlled growth conditions and the numbers and sizes of simulated bacterial dots to create an entire alphabet based on how the colonies would look after they fill a virtual Petri dish. They call this encoding scheme emorfi. The encoding is not one-to-one, as the final simulated pattern corresponding to each letter is not exactly the same every time. However, the researchers discovered that a machine learning program could learn to distinguish between them to recognize the letter intended.

We consider a bandit problem where at any time, the decision maker can add new arms to her consideration set. A new arm is queried at a cost from an "arm-reservoir" containing finitely many "arm-types," each characterized by a distinct mean reward. The cost of query reflects in a diminishing probability of the returned arm being optimal, unbeknown to the decision maker; this feature encapsulates defining characteristics of a broad class of operations-inspired online learning problems, e.g., those arising in markets with churn, or those involving allocations subject to costly resource acquisition. The decision maker's goal is to maximize her cumulative expected payoffs over a sequence of n pulls, oblivious to the statistical properties as well as types of the queried arms. We study two natural modes of endogeneity in the reservoir distribution, and characterize a necessary condition for achievability of sub-linear regret in the problem. We also discuss a UCB-inspired adaptive algorithm that is long-run-average optimal whenever said condition is satisfied, thereby establishing its tightness.

From navigation systems to smart assistants, we communicate with various AI on a daily basis. At the core of such human-AI communication, we convey our understanding of the AI's capability to the AI through utterances with different complexities, and the AI conveys its understanding of our needs and goals to us through system outputs. However, this communication process is prone to failures for two reasons: the AI might have the wrong understanding of the user and the user might have the wrong understanding of the AI. To enhance mutual understanding in human-AI communication, we posit the Mutual Theory of Mind (MToM) framework, inspired by our basic human capability of "Theory of Mind." In this paper, we discuss the motivation of the MToM framework and its three key components that continuously shape the mutual understanding during three stages of human-AI communication. We then describe a case study inspired by the MToM framework to demonstrate the power of MToM framework to guide the design and understanding of human-AI communication.

In this paper we describe the design and the ideas motivating a new Continual Learning benchmark for Autonomous Driving (CLAD), that focuses on the problems of object classification and object detection. The benchmark utilises SODA10M, a recently released large-scale dataset that concerns autonomous driving related problems. First, we review and discuss existing continual learning benchmarks, how they are related, and show that most are extreme cases of continual learning. To this end, we survey the benchmarks used in continual learning papers at three highly ranked computer vision conferences. Next, we introduce CLAD-C, an online classification benchmark realised through a chronological data stream that poses both class and domain incremental challenges; and CLAD-D, a domain incremental continual object detection benchmark. We examine the inherent difficulties and challenges posed by the benchmark, through a survey of the techniques and methods used by the top-3 participants in a CLAD-challenge workshop at ICCV 2021. We conclude with possible pathways to improve the current continual learning state of the art, and which directions we deem promising for future research.

Deep Q-Networks algorithm (DQN) was the first reinforcement learning algorithm using deep neural network to successfully surpass human level performance in a number of Atari learning environments. However, divergent and unstable behaviour have been long standing issues in DQNs. The unstable behaviour is often characterised by overestimation in the $Q$-values, commonly referred to as the overestimation bias. To address the overestimation bias and the divergent behaviour, a number of heuristic extensions have been proposed. Notably, multi-step updates have been shown to drastically reduce unstable behaviour while improving agent's training performance. However, agents are often highly sensitive to the selection of the multi-step update horizon ($n$), and our empirical experiments show that a poorly chosen static value for $n$ can in many cases lead to worse performance than single-step DQN. Inspired by the success of $n$-step DQN and the effects that multi-step updates have on overestimation bias, this paper proposes a new algorithm that we call `Elastic Step DQN' (ES-DQN). It dynamically varies the step size horizon in multi-step updates based on the similarity of states visited. Our empirical evaluation shows that ES-DQN out-performs $n$-step with fixed $n$ updates, Double DQN and Average DQN in several OpenAI Gym environments while at the same time alleviating the overestimation bias.

The goal of lifelong learning is to continuously learn from non-stationary distributions, where the non-stationarity is typically imposed by a sequence of distinct tasks. Prior works have mostly considered idealistic settings, where the identity of tasks is known at least at training. In this paper we focus on a fundamentally harder, so-called task-agnostic setting where the task identities are not known and the learning machine needs to infer them from the observations. Our algorithm, which we call TAME (Task-Agnostic continual learning using Multiple Experts), automatically detects the shift in data distributions and switches between task expert networks in an online manner. At training, the strategy for switching between tasks hinges on an extremely simple observation that for each new coming task there occurs a statistically-significant deviation in the value of the loss function that marks the onset of this new task. At inference, the switching between experts is governed by the selector network that forwards the test sample to its relevant expert network. The selector network is trained on a small subset of data drawn uniformly at random. We control the growth of the task expert networks as well as selector network by employing online pruning. Our experimental results show the efficacy of our approach on benchmark continual learning data sets, outperforming the previous task-agnostic methods and even the techniques that admit task identities at both training and testing, while at the same time using a comparable model size.

Large language models (LLMs) can perform complex reasoning by generating intermediate reasoning steps. Providing these steps for prompting demonstrations is called chain-of-thought (CoT) prompting. CoT prompting has two major paradigms. One leverages a simple prompt like "Let's think step by step" to facilitate step-by-step thinking before answering a question. The other uses a few manual demonstrations one by one, each composed of a question and a reasoning chain that leads to an answer. The superior performance of the second paradigm hinges on the hand-crafting of task-specific demonstrations one by one. We show that such manual efforts may be eliminated by leveraging LLMs with the "Let's think step by step" prompt to generate reasoning chains for demonstrations one by one, i.e., let's think not just step by step, but also one by one. However, these generated chains often come with mistakes. To mitigate the effect of such mistakes, we find that diversity matters for automatically constructing demonstrations. We propose an automatic CoT prompting method: Auto-CoT. It samples questions with diversity and generates reasoning chains to construct demonstrations. On ten public benchmark reasoning tasks with GPT-3, Auto-CoT consistently matches or exceeds the performance of the CoT paradigm that requires manual designs of demonstrations. Code is available at https://github.com/amazon-research/auto-cot

Scheduling plays an important role in automated production. Its impact can be found in various fields such as the manufacturing industry, the service industry and the technology industry. A scheduling problem (NP-hard) is a task of finding a sequence of job assignments on a given set of machines with the goal of optimizing the objective defined. Methods such as Operation Research, Dispatching Rules, and Combinatorial Optimization have been applied to scheduling problems but no solution guarantees to find the optimal solution. The recent development of Reinforcement Learning has shown success in sequential decision-making problems. This research presents a Reinforcement Learning approach for scheduling problems. In particular, this study delivers an OpenAI gym environment with search-space reduction for Job Shop Scheduling Problems and provides a heuristic-guided Q-Learning solution with state-of-the-art performance for Multi-agent Flexible Job Shop Problems.