Recent advancements in feature representation and dimension reduction have highlighted their crucial role in enhancing the efficacy of predictive modeling. This work introduces TemporalPaD, a novel end-to-end deep learning framework designed for temporal pattern datasets. TemporalPaD integrates reinforcement learning (RL) with neural networks to achieve concurrent feature representation and feature reduction. The framework consists of three cooperative modules: a Policy Module, a Representation Module, and a Classification Module, structured based on the Actor-Critic (AC) framework. The Policy Module, responsible for dimensionality reduction through RL, functions as the actor, while the Representation Module for feature extraction and the Classification Module collectively serve as the critic. We comprehensively evaluate TemporalPaD using 29 UCI datasets, a well-known benchmark for validating feature reduction algorithms, through 10 independent tests and 10-fold cross-validation. Additionally, given that TemporalPaD is specifically designed for time series data, we apply it to a real-world DNA classification problem involving enhancer category and enhancer strength. The results demonstrate that TemporalPaD is an efficient and effective framework for achieving feature reduction, applicable to both structured data and sequence datasets.

Aerial Remote Sensing (ARS) vision tasks pose significant challenges due to the unique characteristics of their viewing angles. Existing research has primarily focused on algorithms for specific tasks, which have limited applicability in a broad range of ARS vision applications. This paper proposes the RingMo-Aerial model, aiming to fill the gap in foundation model research in the field of ARS vision. By introducing the Frequency-Enhanced Multi-Head Self-Attention (FE-MSA) mechanism and an affine transformation-based contrastive learning pre-training method, the model's detection capability for small targets is enhanced and optimized for the tilted viewing angles characteristic of ARS. Furthermore, the ARS-Adapter, an efficient parameter fine-tuning method, is proposed to improve the model's adaptability and effectiveness in various ARS vision tasks. Experimental results demonstrate that RingMo-Aerial achieves SOTA performance on multiple downstream tasks. This indicates the practicality and effectiveness of RingMo-Aerial in enhancing the performance of ARS vision tasks.

Transforming a sentence into a two-dimensional (2D) representation (e.g., the table filling) has the ability to unfold a semantic plane, where an element of the plane is a word-pair representation of a sentence which may denote a possible relation representation composed of two named entities. The 2D representation is effective in resolving overlapped relation instances. However, in related works, the representation is directly transformed from a raw input. It is weak to utilize prior knowledge, which is important to support the relation extraction task. In this paper, we propose a two-dimensional feature engineering method in the 2D sentence representation for relation extraction. Our proposed method is evaluated on three public datasets (ACE05 Chinese, ACE05 English, and SanWen) and achieves the state-of-the-art performance. The results indicate that two-dimensional feature engineering can take advantage of a two-dimensional sentence representation and make full use of prior knowledge in traditional feature engineering. Our code is publicly available at https://github.com/Wang-ck123/A-Two-Dimensional-Feature-Engineering-Method-for-Entity-Relation-Extraction

AGI, while powerful, could have negative consequences, warned Diveplane CEO Mike Capps and Liberty Blockchain CCO Christopher Alexander. Legislation moving through the House would provide millions of dollars for research on how to incorporate artificial intelligence into drone technology in an effort to keep the U.S. ahead of China in this increasingly important component of national security. The House Committee on Science, Space, and Technology last week approved legislation from committee Chairman Frank Lucas, R-Okla., that he says needs to pass before China becomes locked in as the world's major supplier of drones. His bill, the National Drone and Advanced Air Mobility Research and Development Act, would fund about $1.6 billion in research over the next five years to give a boost to U.S.-based drone manufacturers. "To say China has cornered this market is an understatement," Lucas said last week. "One single company with extensive ties to the Chinese Communist Party and the People's Liberation Army produces 80% of the drones used recreationally in the U.S." A staff member works on an unmanned aerial vehicle at Guizhou University in Guiyang, China, on May 23, 2023.

Automatic and periodic recompiling of building databases with up-to-date high-resolution images has become a critical requirement for rapidly developing urban environments. However, the architecture of most existing approaches for change extraction attempts to learn features related to changes but ignores objectives related to buildings. This inevitably leads to the generation of significant pseudo-changes, due to factors such as seasonal changes in images and the inclination of building fa\c{c}ades. To alleviate the above-mentioned problems, we developed a contrastive learning approach by validating historical building footprints against single up-to-date remotely sensed images. This contrastive learning strategy allowed us to inject the semantics of buildings into a pipeline for the detection of changes, which is achieved by increasing the distinguishability of features of buildings from those of non-buildings. In addition, to reduce the effects of inconsistencies between historical building polygons and buildings in up-to-date images, we employed a deformable convolutional neural network to learn offsets intuitively. In summary, we formulated a multi-branch building extraction method that identifies newly constructed and removed buildings, respectively. To validate our method, we conducted comparative experiments using the public Wuhan University building change detection dataset and a more practical dataset named SI-BU that we established. Our method achieved F1 scores of 93.99% and 70.74% on the above datasets, respectively. Moreover, when the data of the public dataset were divided in the same manner as in previous related studies, our method achieved an F1 score of 94.63%, which surpasses that of the state-of-the-art method.

He weighs a little over 100 pounds, has auburn dyed hair, and wears a hoodie emblazoned with the words "Dope Shit." But in China's rise to superpower status, the 20-year-old student is a foot soldier playing a vital role. Yin works at Bainiaohe Digital Town, a tech hub set deep in the tree-covered hills of China's southwestern province of Guizhou. The region is traditionally known for growing tea and producing fiery Moutai liquor, but today it's luring Yin, and hundreds of young Chinese like him, to work in the booming sector of Artificial Intelligence or AI. For eight hours each day, they sit at computer terminals on brightly colored swivel chairs and help refine the reliability of facial and voice recognition software, driverless car programs, and even mobile apps that are used to identify plants and insects.

Here in 2018, big data is a big deal in China, and nowhere is this truer than in Guizhou, a remote, impoverished province in southwestern China where the provincial government is trying to build a big data industry from scratch. As He Yuan -- a manager at the Shanghai-based company Beige Big Data, which has an office in Guizhou -- put it to me in a recent visit to the province: "Everyone wants a piece of big data … Many still haven't figured out what the term means, exactly." Despite this lingering confusion, the Chinese government seems fully invested in what The New York Times columnist David Brooks refers to as "data-ism" -- the belief that "everything that can be measured should be measured; that data is a transparent and reliable lens that allows us to filter out emotionalism and ideology; that data will help us do remarkable things -- like foretell the future." And Guizhou -- a province less commonly associated with cutting-edge technology and more often with rugged mountains, poor soil, and extreme poverty -- is trying to position itself at the forefront of this nationwide push. Yet for all its leaders' grand ambitions, several hurdles remain to be overcome.

An exhibition of an unattended vending car built by PIX, a start-up based in Guiyang, Southwest China's Guizhou Province Photo: Courtesy of PIX Imagine waking up and making a reservation for a robot-staffed vending car first thing in the morning so you can enjoy a cup of coffee on the way to work. After work, you get some exercise on the way home via a moving gym. Then a mobile grocery store stops right in front of your door when you arrive home. You select fresh vegetables and ingredients to cook a delicious dinner. PIX claims it will be the first in China to make this convenient future possible.

This week we saw a huge milestone with China beating Amazon to successfully establish fully commercialized drone delivery, a new type of online to offline buying experience combining China's Spotify with retail, and artificial intelligence (AI) predicting which roads will flood in advance in order to reduce traffic congestion. China is pushing the boundaries of technological advancement faster than any country on Earth and This Week In China Tech is the place to stay on top of the news that you won't find in the Western media. Aerial photo taken on May 10, 2018 shows a drone carried with parcels, taking off from a branch post office in Weicheng Township, Qingzhen City of southwest China's Guizhou Province. You probably remember how Amazon captured the consumer imagination when they announced their concept for delivery drones at the end of 2016, but they never really materialized. Chinese retailers have now delivered fully operational drone delivery systems, confirming 17 authorized routes last week (article in Chinese).

A total of 199 deals worth more than 35 billion yuan (5.4 billion U.S. dollars) were signed during the China International Big Data Industry Expo in Guiyang, capital of southwest China's Guizhou Province. Eight panel discussions and 65 forums focusing on artificial intelligence (AI), data security, Internet of Things, shared economy, and targeted poverty alleviation were held during the expo that started on Saturday. Altogether 388 enterprises, more than 50,000 representatives and guests participated in the four-day event, said Chen Yan, mayor of Guiyang. During the expo, Guizhou Province has also published more than 100 projects awaiting for investment, involving the electronic information and application of big data in industries such as service and agriculture. The projects have a combined investment scale of more than 168 billion yuan.