Data Mining Practical Machine Learning Tools and Techniques

Encyclopedia of Complexity and Systems Science

여성사 쓰기의 가능성과 현주소, 한국여성연구소 여성사연구실 지음, 청년사 1999

【臨床医学の展望 2012】 呼吸器病学

Predicting the next location a user tends to visit is an important task for applications like location-based advertising, traffic planning, and tour recommendation. We consider the next location prediction problem for semantic trajectory data, wherein each GPS record is attached with a text message that describes the user's activity. In semantic trajectories, the confluence of spatiotemporal transitions and textual messages indicates user intents at a fine granularity and has great potential in improving location prediction accuracies. Nevertheless, existing methods designed for GPS trajectories fall short in capturing latent user intents for such semantics-enriched trajectory data. We propose a method named semantics-enriched recurrent model (SERM). SERM jointly learns the embeddings of multiple factors (user, location, time, keyword) and the transition parameters of a recurrent neural network in a unified framework. Therefore, it effectively captures semantics-aware spatiotemporal transition regularities to improve location prediction accuracies. Our experiments on two real-life semantic trajectory datasets show that SERM achieves significant improvements over state-of-the-art methods.

SERM: A Recurrent Model for Next Location Prediction in Semantic Trajectories

Medicinal plants are an important element of indigenous medical systems in Viet Nam. These resources are usually regarded as a part of culture's traditional knowledge. One of the prerequisites for any medical recommendation systems and/or applications is accurate identification and classification of medicinal plants. Hence, leveraging technology in automatic classification of these curative herbs has become essential. Unfortunately, building and training a machine learning model from scratch is next to impossible due to the lack of hardware infrastructure and finance support. It painfully restricts the requirements of rapid solutions to deal with the demand. For this purpose, this paper exploits the idea of transfer learning which is the improvement of learning in a new prediction task through the transferability of knowledge from a related prediction task that has already been learned. By utilizing state-of-the-art deep networks re-trained with our collected data, our extensive experiments show that the proposed combination performs perfectly and achieves the classification accuracy of 98.7% within the acceptable training time.

A Combination of Transfer Learning and Deep Learning for Medicinal Plant Classification

Cross-institutional sharing of medical data is essential to provide e.ective collaborative treatment and clinical decisions for patients. Medical data privacy involves ensuring only authorized parties may access the health records under the awareness and approval of patients in any circumstances. This is crucial to any healthcare system because the protection of patients' clinical data is not only an ethical responsibility but also a legal mandate. Despite the importance of medical data sharing, today's healthcare systems have not provided enough protection of patients' sensitive information to be utilized deliberately or unintentionally. Hence, there is an urgent demand for a clinical transaction mechanism that allows patients to access, trace and control their health records. In this paper, the authors focus on several limitations in the literature and propose appropriate improvement in healthcare systems by (i) addressing information security and privacy, (ii) solving the lack of trust between providers, and (iii) encouraging scalability of healthcare interoperability. Building upon these key insights, we introduce several components of a patient-centered healthcare system using smart contracts via blockchain technology. A complete code solution is publicized on the authors' GitHub repository to engage further reproducibility and improvement.

Smart Care: Integrating Blockchain Technology into the Design of Patient-centered Healthcare Systems

The adoption of decentralized cryptocurrency plat-forms is growing fast, thanks to the implementation of Blockchain technology and smart contracts. It encourages the novel frame-works in a wide range of applications including finance and payment methods such as cash on delivery. However, a large number of smart contracts developed for cash on delivery suffer from fraudulent transactions which enable malicious participants to break the signed contracts without sufficient penalties. A shipper will involve in the system and place a mortgage to ensure reliability. A buyer also pledges an amount of money when making the order. Our process not only ensures the interests of a seller but also prevents a fraud shipper. The penalties will be made in two scenarios: (i) the buyer refuses to receive the commodities without any reliable reasons; and (ii) the shipper attempts to make any modification on the delivered goods during transportation. To help developers create more secure and reliable cash on delivery system, we introduce double smart contracts, a framework rooted in Blockchain technology and Ethereum, to tackle those mentioned problems. We also contribute our solution as an open source software that developers can easily add to their implementation to enhance functionality.

/pdf/assuring-non-fraudulent-transactions-in-cash-on-delivery-by-4v5dsqmfc2.pdf

Assuring Non-fraudulent Transactions in Cash on Delivery by Introducing Double Smart Contracts

Previous research on content-based geolocation in general has developed prediction methods via conducting pre-partitioning and applying classification methods. The input of these methods is the concatenation of individual tweets during a period of time. But unfortunately, these methods have some drawbacks. They discard the natural real-values properties of latitude and longitude as well as fail to capture geolocation in near real-time. In this work, we develop a novel generative content-based regression model via a matrix factorization technique to tackle the near real-time geolocation prediction problem. With this model, we aim to address a couple of un-answered questions. First, we prove that near real-time geolocation prediction can be accomplished if we leave out the concatenation. Second, we account the real-values properties of physical coordinates within a regression solution. We apply our model on Twitter datasets as an example to prove the effectiveness and generality. Our experimental results show that the proposed model, in the best scenario, outperforms a set of state-of-the-art regression models including Support Vector Machines and Factorization Machines by a reduction of the median localization error up to 79%.

Near Real-time Geolocation Prediction in Twitter Streams via Matrix Factorization Based Regression

In this paper, we propose a patient-centric care system based on a smart contract mechanism which is an enhancement to Blockchain technologies. The health records are controlled by the patients, it means the patients could choose and share what kind of information related to their health history to the clinics even the sensitive information. The secure information mechanism and e.ective technical infrastructures to enhance care coordination of individuals and communities are achieved by using Smart Contract which is implemented on Hyperledger Fabric and Ethereum Blockchain. Six algorithms were utilized for interacting data between the components of a Patient-centered healthcare system: createData(), query-Data(), get.eryHistory(), createPatientInformation() / createMedicalRecord() / createDrugInformation() / createHospitalFees() and modifyPatientInformation() / modifyMedicalRecord() / modifyDrug-Information() / modifyHospitalFees(). The simulation results show that the healthcare system and record resources consumption by the aforementioned six algorithms, in terms of execution speed and memory capacity that are allocated for each function at the execution time ranges from 1s.

Nghia Duong-Trung

Papers

A Combination of Transfer Learning and Deep Learning for Medicinal Plant Classification

Smart Care: Integrating Blockchain Technology into the Design of Patient-centered Healthcare Systems

Assuring Non-fraudulent Transactions in Cash on Delivery by Introducing Double Smart Contracts

Near Real-time Geolocation Prediction in Twitter Streams via Matrix Factorization Based Regression

On Components of a Patient-centered Healthcare System Using Smart Contract