We started as a small telegram group consisting of some old friends working on some shady businesses (lol. Personally, I don't know when is the exact creation date of our telegram group named "Anti Wacana", means "no-nonsense" and, by definition, is a group of people that always act without non-sense. Officially, we said that our founding date was on August 2020 without knowing the exact day. Most of us are working full-time as Data Scientists in different companies while doing some projects in "Anti Wacana". We handle projects in many areas not limited to Data Science but also in Computer Vision, Software Engineering, web, and mobile app development stuff.

Thompson sampling and other Bayesian sequential decision-making algorithms are among the most popular approaches to tackle explore/exploit trade-offs in (contextual) bandits. The choice of prior in these algorithms offers flexibility to encode domain knowledge but can also lead to poor performance when misspecified. In this paper, we demonstrate that performance degrades gracefully with misspecification. We prove that the expected reward accrued by Thompson sampling (TS) with a misspecified prior differs by at most $\tilde{\mathcal{O}}(H^2 \epsilon)$ from TS with a well specified prior, where $\epsilon$ is the total-variation distance between priors and $H$ is the learning horizon. Our bound does not require the prior to have any parametric form. For priors with bounded support, our bound is independent of the cardinality or structure of the action space, and we show that it is tight up to universal constants in the worst case. Building on our sensitivity analysis, we establish generic PAC guarantees for algorithms in the recently studied Bayesian meta-learning setting and derive corollaries for various families of priors. Our results generalize along two axes: (1) they apply to a broader family of Bayesian decision-making algorithms, including a Monte-Carlo implementation of the knowledge gradient algorithm (KG), and (2) they apply to Bayesian POMDPs, the most general Bayesian decision-making setting, encompassing contextual bandits as a special case. Through numerical simulations, we illustrate how prior misspecification and the deployment of one-step look-ahead (as in KG) can impact the convergence of meta-learning in multi-armed and contextual bandits with structured and correlated priors.

The incompleteness of Knowledge Graphs (KGs) is a crucial issue affecting the quality of AI-based services. In the scholarly domain, KGs describing research publications typically lack important information, hindering our ability to analyse and predict research dynamics. In recent years, link prediction approaches based on Knowledge Graph Embedding models became the first aid for this issue. In this work, we present Trans4E, a novel embedding model that is particularly fit for KGs which include N to M relations with N$\gg$M. This is typical for KGs that categorize a large number of entities (e.g., research articles, patents, persons) according to a relatively small set of categories. Trans4E was applied on two large-scale knowledge graphs, the Academia/Industry DynAmics (AIDA) and Microsoft Academic Graph (MAG), for completing the information about Fields of Study (e.g., 'neural networks', 'machine learning', 'artificial intelligence'), and affiliation types (e.g., 'education', 'company', 'government'), improving the scope and accuracy of the resulting data. We evaluated our approach against alternative solutions on AIDA, MAG, and four other benchmarks (FB15k, FB15k-237, WN18, and WN18RR). Trans4E outperforms the other models when using low embedding dimensions and obtains competitive results in high dimensions.

Intelligence necessitates memory. Without memory, humans fail to perform various nontrivial tasks such as reading novels, playing games or solving maths. As the ultimate goal of machine learning is to derive intelligent systems that learn and act automatically just like human, memory construction for machine is inevitable. Artificial neural networks model neurons and synapses in the brain by interconnecting computational units via weights, which is a typical class of machine learning algorithms that resembles memory structure. Their descendants with more complicated modeling techniques (a.k.a deep learning) have been successfully applied to many practical problems and demonstrated the importance of memory in the learning process of machinery systems. Recent progresses on modeling memory in deep learning have revolved around external memory constructions, which are highly inspired by computational Turing models and biological neuronal systems. Attention mechanisms are derived to support acquisition and retention operations on the external memory. Despite the lack of theoretical foundations, these approaches have shown promises to help machinery systems reach a higher level of intelligence. The aim of this thesis is to advance the understanding on memory and attention in deep learning. Its contributions include: (i) presenting a collection of taxonomies for memory, (ii) constructing new memory-augmented neural networks (MANNs) that support multiple control and memory units, (iii) introducing variability via memory in sequential generative models, (iv) searching for optimal writing operations to maximise the memorisation capacity in slot-based memory networks, and (v) simulating the Universal Turing Machine via Neural Stored-program Memory-a new kind of external memory for neural networks.

In spite of the great success of deep learning technologies, training Imagine a realtime AI-as-a-Service (AIaaS) system that can instantly and delivery of a practically serviceable model is still a highly deliver resource-efficient models for any on-demand tasks to multiple time-consuming process. Furthermore, a resulting model is usually users who can be non-expertise in AI. For example, a mobile too generic and heavyweight, and hence essentially goes through user should quickly adapt to a dynamically changing environment another expensive model compression phase to fit in a resourcelimited (e.g., entering a restaurant in an animal theme park and returning to device like embedded systems. Inspired by the fact that see animals having lunch) preferably without continually training a machine learning task specifically requested by mobile users is a large generic model due to the resource constraints. In this case, often much simpler than it is supported by a massive generic model, such a realtime AIaaS system would be an adequate solution so that this paper proposes a framework, called Pool of Experts (PoE), that a user can immediately be given a properly functioning model by instantly builds a lightweight and task-specific model without any the system without any training overhead. Although AIaaS is getting training process. For a realtime model querying service, PoE first popular to the point that global vendors have initiated code-free extracts a pool of primitive components, called experts, from a machine learning platforms, it still takes a while to train a specific well-trained and sufficiently generic network by exploiting a novel neural network that achieves a practical level of the accuracy even conditional knowledge distillation method, and then performs our with the help of transfer learning from a massive pretrained model, train-free knowledge consolidation to quickly combine necessary often referred to as oracle. To realize a realtime AIaaS system that experts into a lightweight network for a target task. Thanks to this immediately builds a task-specific and lightweight model, would it train-free property, in our thorough empirical study, PoE can build be possible to preprocess such an oracle neural network so that any a fairly accurate yet compact model in a realtime manner, whereas queried knowledge is efficiently extracted with no training at all? it takes a few minutes per query for the other training methods to The fundamental problem of extracting knowledge from a pretrained achieve a similar level of the accuracy.

Spatial joins allow to augment one spatial dataset with information from another spatial dataset by linking overlapping features. In this post I will provide an example showing how to augment a dataset containing school locations with socioeconomic data of their surrounding statistical region using R and the package sf (Pebesma 2018). This approach has the drawback that the surrounding statistical region doesn't reflect the actual catchment area of the school. I will present an alternative approach where the overlaps of the schools' catchment areas with the statistical regions allow to calculate the weighted average of the socioeconomic statistics. If we have no data about the actual catchment areas of the schools, we may resort to approximating these areas as circular regions or as Voronoi regions around schools.

The analytics leader of a US-based Fortune 200 company was under severe pressure. Her team supported 45,000 employees of the global energy company, and the business users weren't happy. The analytics deliverables were often late and suffered from poor quality. The analytics team was a part of the IT organization and was struggling to fill their open positions. The skills needed couldn't be found within the IT team.

Systemic bias with respect to gender, race and ethnicity, often unconscious, is prevalent in datasets involving choices among individuals. Consequently, society has found it challenging to alleviate bias and achieve diversity in a way that maintains meritocracy in such settings. We propose (a) a novel optimization approach based on optimally flipping outcome labels and training classification models simultaneously to discover changes to be made in the selection process so as to achieve diversity without significantly affecting meritocracy, and (b) a novel implementation tool employing optimal classification trees to provide insights on which attributes of individuals lead to flipping of their labels, and to help make changes in the current selection processes in a manner understandable by human decision makers. We present case studies on three real-world datasets consisting of parole, admissions to the bar and lending decisions, and demonstrate that the price of diversity is low and sometimes negative, that is we can modify our selection processes in a way that enhances diversity without affecting meritocracy significantly, and sometimes improving it.

Inducing causal relationships from observations is a classic problem in machine learning. Most work in causality starts from the premise that the causal variables themselves are observed. However, for AI agents such as robots trying to make sense of their environment, the only observables are low-level variables like pixels in images. To generalize well, an agent must induce high-level variables, particularly those which are causal or are affected by causal variables. A central goal for AI and causality is thus the joint discovery of abstract representations and causal structure. However, we note that existing environments for studying causal induction are poorly suited for this objective because they have complicated task-specific causal graphs which are impossible to manipulate parametrically (e.g., number of nodes, sparsity, causal chain length, etc.). In this work, our goal is to facilitate research in learning representations of high-level variables as well as causal structures among them. In order to systematically probe the ability of methods to identify these variables and structures, we design a suite of benchmarking RL environments. We evaluate various representation learning algorithms from the literature and find that explicitly incorporating structure and modularity in models can help causal induction in model-based reinforcement learning.

In recent years, machine learning has begun automating decision making in fields as varied as college admissions, credit lending, and criminal sentencing. The socially sensitive nature of some of these applications together with increasing regulatory constraints has necessitated the need for algorithms that are both fair and interpretable. In this paper we consider the problem of building Boolean rule sets in disjunctive normal form (DNF), an interpretable model for binary classification, subject to fairness constraints. We formulate the problem as an integer program that maximizes classification accuracy with explicit constraints on two different measures of classification parity: equality of opportunity and equalized odds. Column generation framework, with a novel formulation, is used to efficiently search over exponentially many possible rules. When combined with faster heuristics, our method can deal with large data-sets. Compared to other fair and interpretable classifiers, our method is able to find rule sets that meet stricter notions of fairness with a modest trade-off in accuracy.