scholarly journals Diversity-sensitive social platforms and responsibility

2021 ◽  
Vol 21 (2) ◽  
pp. 43
Author(s):  
Karoline Reinhardt
Keyword(s):  
Machine Learning ◽  
Social Interaction ◽  
Personal Data ◽  
Ethical Framework ◽  
Human Diversity ◽  
Ongoing Debate ◽  
Related Data ◽  
Interaction Protocols ◽  
Time Diversity ◽  
Computer Scientists

There is an ongoing debate on how algorithms and machine learning can and should deal with human diversity while avoiding the pitfalls of statistical stereotyping, the re-enforcement of clichés and the perpetuation of unjust discrimination. Computer scientists try to tackle these issues by developing algorithms and social-interaction protocols for mediating diversity-aware interactions between people, for instance on diversity-sensitive social platforms. At the same time, diversity-related data often comprise sensitive personal data, and their collection, storage and management increases the vulnerability of users to various misuse scenarios. Already this observation leads to the question, how do we need to conceptualize responsibility to do justice to the increased vulnerability? In this paper, I thus focus on the questions a diversity-sensitive social platform raises with regard to responsibility, and propose a tentative ethical framework of responsibility for these platforms.

Deep Learning in Disease Diagnosis: Models and Datasets

2020 ◽  
Vol 15 ◽  
Author(s):  
Deeksha Saxena ◽  
Mohammed Haris Siddiqui ◽  
Rajnish Kumar
Keyword(s):  
Biological Sciences ◽  
Machine Learning ◽  
Deep Learning ◽  
Disease Diagnosis ◽  
Learning Models ◽  
Data Types ◽  
Related Data ◽  
Abstract Level ◽  
Experimental Validations ◽  
Selection Of

Background: Deep learning (DL) is an Artificial neural network-driven framework with multiple levels of representation for which non-linear modules combined in such a way that the levels of representation can be enhanced from lower to a much abstract level. Though DL is used widely in almost every field, it has largely brought a breakthrough in biological sciences as it is used in disease diagnosis and clinical trials. DL can be clubbed with machine learning, but at times both are used individually as well. DL seems to be a better platform than machine learning as the former does not require an intermediate feature extraction and works well with larger datasets. DL is one of the most discussed fields among the scientists and researchers these days for diagnosing and solving various biological problems. However, deep learning models need some improvisation and experimental validations to be more productive. Objective: To review the available DL models and datasets that are used in disease diagnosis. Methods: Available DL models and their applications in disease diagnosis were reviewed discussed and tabulated. Types of datasets and some of the popular disease related data sources for DL were highlighted. Results: We have analyzed the frequently used DL methods, data types and discussed some of the recent deep learning models used for solving different biological problems. Conclusion: The review presents useful insights about DL methods, data types, selection of DL models for the disease diagnosis.

scholarly journals An Annotated Corpus of Crime-Related Portuguese Documents for NLP and Machine Learning Processing

Data ◽  
2021 ◽  
Vol 6 (7) ◽  
pp. 71
Author(s):  
Gonçalo Carnaz ◽  
Mário Antunes ◽  
Vitor Beires Nogueira
Keyword(s):  
Machine Learning ◽  
Language Processing ◽  
Named Entity Recognition ◽  
Entity Recognition ◽  
Automatic Identification ◽  
Named Entities ◽  
Related Data ◽  
Named Entity ◽  
Chain Of Custody ◽  
Evidence Collection

Criminal investigations collect and analyze the facts related to a crime, from which the investigators can deduce evidence to be used in court. It is a multidisciplinary and applied science, which includes interviews, interrogations, evidence collection, preservation of the chain of custody, and other methods and techniques of investigation. These techniques produce both digital and paper documents that have to be carefully analyzed to identify correlations and interactions among suspects, places, license plates, and other entities that are mentioned in the investigation. The computerized processing of these documents is a helping hand to the criminal investigation, as it allows the automatic identification of entities and their relations, being some of which difficult to identify manually. There exists a wide set of dedicated tools, but they have a major limitation: they are unable to process criminal reports in the Portuguese language, as an annotated corpus for that purpose does not exist. This paper presents an annotated corpus, composed of a collection of anonymized crime-related documents, which were extracted from official and open sources. The dataset was produced as the result of an exploratory initiative to collect crime-related data from websites and conditioned-access police reports. The dataset was evaluated and a mean precision of 0.808, recall of 0.722, and F1-score of 0.733 were obtained with the classification of the annotated named-entities present in the crime-related documents. This corpus can be employed to benchmark Machine Learning (ML) and Natural Language Processing (NLP) methods and tools to detect and correlate entities in the documents. Some examples are sentence detection, named-entity recognition, and identification of terms related to the criminal domain.

scholarly journals Federated Learning in a Medical Context: A Systematic Literature Review

2021 ◽  
Vol 21 (2) ◽  
pp. 1-31
Author(s):  
Bjarne Pfitzner ◽  
Nico Steckhan ◽  
Bert Arnrich
Keyword(s):  
Machine Learning ◽  
Literature Review ◽  
Systematic Literature Review ◽  
Data Privacy ◽  
Research Area ◽  
Learning Models ◽  
Related Data ◽  
Private Data ◽  
Large Databases ◽  
Machine Learning Models

Data privacy is a very important issue. Especially in fields like medicine, it is paramount to abide by the existing privacy regulations to preserve patients’ anonymity. However, data is required for research and training machine learning models that could help gain insight into complex correlations or personalised treatments that may otherwise stay undiscovered. Those models generally scale with the amount of data available, but the current situation often prohibits building large databases across sites. So it would be beneficial to be able to combine similar or related data from different sites all over the world while still preserving data privacy. Federated learning has been proposed as a solution for this, because it relies on the sharing of machine learning models, instead of the raw data itself. That means private data never leaves the site or device it was collected on. Federated learning is an emerging research area, and many domains have been identified for the application of those methods. This systematic literature review provides an extensive look at the concept of and research into federated learning and its applicability for confidential healthcare datasets.

scholarly journals Machine Learning in P&C Insurance: A Review for Pricing and Reserving

Risks ◽  
2020 ◽  
Vol 9 (1) ◽  
pp. 4 ◽  
Author(s):  
Christopher Blier-Wong ◽  
Hélène Cossette ◽  
Luc Lamontagne ◽  
Etienne Marceau
Keyword(s):  
Machine Learning ◽  
Insurance Industry ◽  
Learning Algorithms ◽  
Machine Learning Algorithms ◽  
Actuarial Science ◽  
Science Field ◽  
The Past ◽  
Highly Nonlinear ◽  
Computer Scientists ◽  
Applications Of Machine Learning

In the past 25 years, computer scientists and statisticians developed machine learning algorithms capable of modeling highly nonlinear transformations and interactions of input features. While actuaries use GLMs frequently in practice, only in the past few years have they begun studying these newer algorithms to tackle insurance-related tasks. In this work, we aim to review the applications of machine learning to the actuarial science field and present the current state of the art in ratemaking and reserving. We first give an overview of neural networks, then briefly outline applications of machine learning algorithms in actuarial science tasks. Finally, we summarize the future trends of machine learning for the insurance industry.

scholarly journals Use of Existing Health Data in Epidemiologic Research–Issues of Informed Consent Under Normal Circumstances and at a Time of Health Crises

2020 ◽  
pp. 1-4
Author(s):  
Carsten Obel ◽  
Carsten Obel ◽  
Jørn Olsen ◽  
Uffe Juul Jensen
Keyword(s):  
Informed Consent ◽  
Personal Data ◽  
Health Data ◽  
Epidemiologic Research ◽  
Health Crisis ◽  
Research Issues ◽  
Data Analyses ◽  
Computer Scientists ◽  
Patient Groups ◽  
Access To Data

In epidemiologic research we study why we get sick and how we get better. To do this we frequently need large datasets on exposure, diagnoses, treatment and more. We need data often classified as sensitive and regulated by law stating a need for informed consent. We argue that modern epidemiologic research often can be done on existing data without having informed consent and without violating basic ethic principles. We also argue for a timely and fair access to data in approved project. Modern encryption technics and methods of data analyses can reduce the risk of disclosure of personal data to a level close to what we have for anonymous data. If we allow open use of administrative health data and existing research data, we will be able to produce much more information to advance disease prevention, health promotion and treatment. Epidemiologists should collaborate more with computer scientists and patient groups in developing/implementing principles for ‘modern methods of data analyses’. Under a severe health crisis data are in high demand to provide the information needed to prevent deaths and diseases and often time does not permit requiring ‘informed consent’. Such a situation in now plying out worldwide under the Covid-19 pandemic.

scholarly journals Flood Prediction and Warning System using Dam Data Monitoring

2020 ◽  
Vol 9 (9) ◽  
pp. 294-298
Keyword(s):  
Machine Learning ◽  
Flood Hazard ◽  
Warning System ◽  
Support Vector ◽  
Flood Prediction ◽  
Loss Of Life ◽  
Related Data ◽  
Dam Management ◽  
Weather Parameters ◽  
The Impact

Flood is one of the most devastating natural calamities affecting parts of the state from past few years. The recurring calamity necessitates an efficient early warning system since anticipation and preparedness play a key role in mitigating the impact. Though heavy and erratic rainfall has been marked as one of the main reasons for flood in several places, flood witnessed by various regions of Kerala was the result of sudden opening of reservoirs indicating poor dam management. The unforeseen flow of water often provided less time for evacuation. Prediction thus plays key role in avoiding loss of life and property, followed by such calamities. The vast benefits and potentials offered by Machine Learning makes it the most promising approach. The developed system is a model by taking Malampuzha Dam as reference. Support Vector Machine (SVM) is used as machine learning method for prediction and is programmed in python. The idea has been to create early flood prediction and warning system by monitoring different weather parameters and dam-related data. The feature vectors include current live storage, current reservoir level, rainfall and relative humidity from the period 2016-2019. Based on the analysis of these parameters, the open/closure of shutters of the dam is predicted. Release of shutters has varied impacts in the nearby regions and is measured by succeeding prediction, by mapping regions on grounds of level warning to be issued. Warning is issued through Flask-based server, by identifying vulnerable areas based on flood hazard reference for regions. The dam status prediction model delivered highest prediction accuracy of 99.14% and associated levels of warning has been generated in the development server, thus preventing unexpected release.

Applications of Machine Learning in Cyber Security

2020 ◽  
pp. 83-103 ◽  
Author(s):  
Charu Virmani ◽  
Tanu Choudhary ◽  
Anuradha Pillai ◽  
Manisha Rani
Keyword(s):  
Machine Learning ◽  
Cyber Security ◽  
Machine Learning Algorithms ◽  
Technological Systems ◽  
The World ◽  
Computer Scientists ◽  
Potential Risks ◽  
To Come ◽  
Time Basis ◽  
Applications Of Machine Learning

With the exponential rise in technological awareness in the recent decades, technology has taken over our lives for good, but with the application of computer-aided technological systems in various domains of our day-to-day lives, the potential risks and threats have also come to the fore, aiming at the various security features that include confidentiality, integrity, authentication, authorization, and so on. Computer scientists the world over have tried to come up, time and again, with solutions to these impending problems. With time, attackers have played out complicated attacks on systems that are hard to comprehend and even harder to mitigate. The very fact that a huge amount of data is processed each second in organizations gave birth to the concept of Big Data, thereby making the systems more adept and intelligent in dealing with unprecedented attacks on a real-time basis. This chapter presents a study about applications of machine learning algorithms in cyber security.

A Framework for Protecting Users' Privacy in Cloud

2019 ◽  
pp. 378-389
Author(s):  
Adesina S. Sodiya ◽  
Adegbuyi B.
Keyword(s):  
Machine Learning ◽  
Cloud Computing ◽  
Privacy Preservation ◽  
Personal Data ◽  
Decomposition Algorithm ◽  
Privacy Concerns ◽  
Personally Identifiable Information ◽  
Foreign Key ◽  
Huge Challenge

Data and document privacy concerns are increasingly important in the online world. In Cloud Computing, the story is the same, as the secure processing of personal data represents a huge challenge. The main focus is to preserve and protect personally identifiable information (PII) of individuals, customers, businesses, governments and organisations. The current use of anonymization techniques is not quite efficient because of its failure to use the structure of the datasets under consideration and inability to use a metric that balances the usefulness of information with privacy preservation. In this work, an adaptive lossy decomposition algorithm was developed for preserving privacy in cloud computing. The algorithm uses the foreign key associations to determine the generalizations possible for any attribute in the database. It generates penalties for each obscured attribute when sharing and proposes an optimal decomposition of the relation. Postgraduate database of Federal University of Agriculture, Abeokuta, Nigeria and Adult database provided at the UCIrvine Machine Learning Repository were used for the evaluation. The result shows a system that could be used to improve privacy in cloud computing.

scholarly journals A Data Protection Framework for Learning Analytics

2016 ◽  
Vol 3 (1) ◽  
Author(s):  
Andrew Nicholas Cormack
Keyword(s):  
Data Protection ◽  
Learning Analytics ◽  
Personal Data ◽  
Informed Choice ◽  
Ethical Framework ◽  
Early Interventions ◽  
Sensitive Data ◽  
Analysis Pattern ◽  
Open Questions ◽  
Automated Processing

Most studies on the use of digital student data adopt an ethical framework derived from human-studies research, based on the informed consent of the experimental subject. However consent gives universities little guidance on the use of learning analytics as a routine part of educational provision: which purposes are legitimate and which analyses involve an unacceptable risk of harm. Obtaining consent when students join a course will not give them meaningful control over their personal data three or more years later. Relying on consent may exclude those most likely to benefit from early interventions. This paper proposes an alternative framework based on European Data Protection law. Separating the processes of analysis (pattern-finding) and intervention (pattern-matching) gives students and staff continuing protection from inadvertent harm during data analysis; students have a fully informed choice whether or not to accept individual interventions; organisations obtain clear guidance: how to conduct analysis, which analyses should not proceed, and when and how interventions should be offered. The framework provides formal support for practices that are already being adopted and helps with several open questions in learning analytics, including its application to small groups and alumni, automated processing and privacy-sensitive data.

Sign in / Sign up

Export Citation Format

Share Document