Code language models have emerged as useful tools for various programming tasks, yet they often struggle when it comes to complex ones. In this paper, we explore the potential of curriculum learning in enhancing the performance of these models. While prior research has suggested that curriculum learning does not necessarily help in improving the performance of language models, our results surprisingly show that this may not be the case for code language models. We demonstrate that a well-designed curriculum learning approach significantly improves the accuracy of small decoder-only code language models on the task of code execution, while its effect on code completion is less significant. To explore the potential of curriculum learning, we train multiple GPT models with 1 million parameters each to predict the next token and evaluate them on code completion and execution tasks. Our contributions include proposing a novel code difficulty assessment metric by combining software code measures, investigating the effectiveness of Curriculum Learning for code language models, and introducing a Novel Curriculum Learning schedule that enhances the performance of small decoder-only language models in code execution tasks. The results of this paper open the door for more research on the use of curriculum learning for code language models.

The scarcity of high quality actions video data is a bottleneck in the research and application of action recognition. Although significant effort has been made in this area, there still exist gaps in the range of available data types a more flexible and comprehensive data set could help bridge. In this paper, we present a new 3D actions data simulation engine and generate 3 sets of sample data to demonstrate its current functionalities. With the new data generation process, we demonstrate its applications to image classifications, action recognitions and potential to evolve into a system that would allow the exploration of much more complex action recognition tasks. In order to show off these capabilities, we also train and test a list of commonly used models for image recognition to demonstrate the potential applications and capabilities of the data sets and their generation process.

In 2022, the marketing industry is being compelled to develop a new way of thinking: one which recognises and respects consumer concerns around data privacy and the global environment. Earlier this summer, Remi Lémonnier, Co-Founder and President of Scibids coined a new term, 'Virtuous Advertising', to describe the future that marketers should be moving towards. And why does Artificial Intelligence (AI) play a big role in making Virtuous Advertising the new norm? Virtuous Advertising acknowledges that marketing organisations are under increasing pressure due to the heightened economic, environmental and regulatory pressures surrounding digital marketing. At the heart of the concept of Virtuous Advertising is respecting resources, i.e. being efficient so that we only use resources that we absolutely need.

Artificial intelligence or our capability of inventing systems that can think, create and act better and faster than humans has become a buzzword that everyone knows about. AI for medicine, AI for finance, AI for auto, we all seem to know and be fearful for the role of AI in next stage of human and industrial evolution. So, why is AI important in this time and age? Is it because the world is growing so rapidly in terms of our demands and we are not able to keep up with the pace of meeting those demands, as manual operations are inefficient? Or is it getting rid of human errors in execution?

Synthetic population generation is the process of combining multiple socioeonomic and demographic datasets from various sources and at different granularity, and downscaling them to an individual level. Although it is a fundamental step for many data science tasks, an efficient and standard framework is absent. In this study, we propose a multi-stage framework called SynC (Synthetic Population via Gaussian Copula) to fill the gap. SynC first removes potential outliers in the data and then fits the filtered data with a Gaussian copula model to correctly capture dependencies and marginal distributions of sampled survey data. Finally, SynC leverages neural networks to merge datasets into one and then scales them accordingly to match the marginal constraints. We make four key contributions in this work: 1) propose a novel framework for generating individual level data from aggregated data sources by combining state-of-the-art machine learning and statistical techniques, 2) design a metric for validating the accuracy of generated data when the ground truth is hard to obtain, 3) demonstrate its effectiveness with the Canada National Census data and presenting two real-world use cases where datasets of this nature can be leveraged by businesses, and 4) release an easy-to-use framework implementation for reproducibility.

For manufacturers, managing the supply chain from beginning to end has been like traveling two superhighways interrupted by a long stretch of dirt road. Manufacturers have benefited from increasingly powerful tools for demand planning and logistics management – the first and last parts of their supply chains – but tracking the performance of manufacturing production across the supply chain has remained stuck in the era of clipboards, whiteboards, spreadsheets and manually assembled reports. For most companies, understanding machine capacity, throughput, efficiency, and quality across the supply chain remains a black box. Companies that rely heavily on contract manufacturers have even less visibility – challenged by partners with different systems, processes, and levels of willingness to collaborate. Today's supply chain monitoring systems lack the ability to look at machine and part/batch-level data across the supply chain, limiting a global manufacturer's ability to manage their supplier base as an integrated platform.

In my last post I looked at generating synthetic data sets with the'synthpop' package, some of the challenges and neat things the package can do. It is simple to use which is great when you have a single data set with independent features. This post will build on the last post by tackling other complications when attempting to synthesise data. These challenges occur regularly in practice and this post will offer a couple of solutions, but there are plenty more. I'll detail how more complex synthesis can be done using synthpop.

Bidding optimisation is considered among toughest critical problems in online advertising. Bidding strategies adopt different search pattern, for example, Sponsored Search (SS) which depends on the randomness of the user's behaviour and the nature of the platform. Display advertising is considered as one of the simple techniques for auction and has taken over Real-Time Bidding resulting in a better performance for the advertisers. In this article, we will explore how Deep Learning techniques are implemented to optimise the Sponsored Search Real Time Bidding (SS-RTB) system in a stochastic environment. A Reinforcement Learning solution for handling the stochastic environment is proposed in the paper titled Deep Reinforcement Learning For Sponsored Search Real Time Bidding by Alibaba group, where the state transition probability is considered for every two days.

The tight long-run fit of the learning curve has led to its use as a tool to predict the future cost of solar panels. Nemet (2006) is skeptical of the view that learning has been an important driver of cost reduction, and uses data during 1975-2002 to show that increases in plant size has been the most important driver of reduction in cost per watt.