scholarly journals A Data Mining-based Cross-Industry Process for Predicting Major Bleeding in Mechanical Circulatory Support

2021 ◽  
Author(s):  
S E A Felix ◽  
A Bagheri ◽  
F R Ramjankhan ◽  
M R Spruit ◽  
D Oberski ◽  
...  
Keyword(s):  
Data Mining ◽  
Major Bleeding ◽  
Predictive Models ◽  
Mechanical Circulatory Support ◽  
Predictive Analytics ◽  
Area Under The Curve ◽  
Circulatory Support ◽  
Contributing Factors ◽  
Using Data ◽  
Near Future

Abstract Background Over a third of patients, treated with mechanical circulatory support (MCS) for end-stage heart failure, experience major bleeding. Currently, the prediction of a major bleeding in the near future is difficult because of many contributing factors. Objectives Predictive analytics using data mining could help calculating the risk of bleeding, however its application is generally reserved for experienced researchers on this subject. We propose an easy applicable data mining tool to predict major bleeding in MCS patients. Methods All data of electronic health records of MCS patients in the University Medical Centre Utrecht were included. Based on the cross-industry standard process for data mining (CRISP-DM) methodology, an application named Auto-Crisp was developed. Auto-Crisp was used to evaluate the predictive models for a major bleeding in the next 3, 7 and 30 days after the first 30 days postoperatively following MCS implantation. The performance of the predictive models are investigated by the area under the curve (AUC) evaluation measure. Results In 25.6% of 273 patients, a total of 142 major bleedings occurred during a median follow-up period of 542 (IQR 205–1044) days. The best predictive models assessed by Auto-Crisp had AUC values of 0.792, 0.788, and 0.776 for bleedings in the next 3, 7, and 30 days, respectively. Conclusion The Auto-Crisp-based predictive model created in this study had an acceptable performance to predict major bleeding in MCS patients in the near future. However, further validation of the application is needed to evaluate Auto-Crisp in other research projects.

scholarly journals Retail Giant Sales Forecasting using Machine Learning

2019 ◽  
Vol 8 (2S11) ◽  
pp. 2408-2411
Keyword(s):  
Machine Learning ◽  
Data Mining ◽  
Decision Making ◽  
Learning Model ◽  
Accurate Prediction ◽  
Machine Learning Model ◽  
Different Types ◽  
Using Data ◽  
Near Future ◽  
Sales Trend

Sales forecasting is widely recognized and plays a major role in an organization’s decision making. It is an integral part in business execution of retail giants, so that they can change their strategy to improve sales in the near future. This helps in better management of their resources like machine, money and manpower. Forecasting the sales will help in managing the revenue and inventory accordingly. This paper proposes a model that can forecast most profitable segments at granular level. As most retail giants have many branches in different locations, consolidation of sales are hard using data mining. Instead using machine learning model helps in getting reliable and accurate results. This paper helps in understanding the sales trend to monitor or predict future applicable on different types of sales patterns and products to produce accurate prediction results.

A predictive analytics of physicians prescription and pharmacies sales correlation using data mining

2019 ◽  
Vol 13 (3) ◽  
pp. 346-363
Author(s):  
Babak Sohrabi ◽  
Iman Raeesi Vanani ◽  
Nastaran Nikaein ◽  
Saeideh Kakavand
Keyword(s):  
Data Mining ◽  
Decision Making ◽  
Pharmaceutical Industry ◽  
Predictive Analytics ◽  
Customer Relationship ◽  
Sales Managers ◽  
Content Type ◽  
Using Data ◽  
The Impact ◽  
The Relationship

Purpose In the pharmaceutical industry, marketing and sales managers often deal with massive amounts of marketing and sales data. One of their biggest concerns is to recognize the impact of actions taken on sold-out products. Data mining discovers and extracts useful patterns from such large data sets to find hidden and worthy patterns for the decision-making. This paper, too, aims to demonstrate the ability of data-mining process in improving the decision-making quality in the pharmaceutical industry. Design/methodology/approach This research is descriptive in terms of the method applied, as well as the investigation of the existing situation and the use of real data and their description. In fact, the study is quantitative and descriptive, from the point of view of its data type and method. This research is also applicable in terms of purpose. The target population of this research is the data of a pharmaceutical company in Iran. Here, the cross-industry standard process for data mining methodology was used for data mining and data modeling. Findings With the help of different data-mining techniques, the authors could examine the effect of the visit of doctors overlooking the pharmacies and the target was set for medical representatives on the pharmaceutical sales. For that matter, the authors used two types of classification rules: decision tree and neural network. After the modeling of algorithms, it was determined that the two aforementioned rules can perform the classification with high precision. The results of the tree ID3 were analyzed to identify the variables and path of this relationship. Originality/value To the best of the authors’ knowledge, this is one of the first studies to provide the real-world direct empirical evidence of “Analytics of Physicians Prescription and Pharmacies Sales Correlation Using Data Mining.” The results showed that the most influential variables of “the relationship between doctors and their visits to pharmacies,” “the length of customer relationship” and “the relationship between the sale of pharmacies and the target set for medical representatives” were “deviation from the implementation plan.” Therefore, marketing and sales managers must pay special attention to these factors while planning and targeting for representatives. The authors could focus only on a small part of this study.

Selecting Children for Mechanical Circulatory Support

2019 ◽  
pp. 54-60
Author(s):  
Devon O. Aganga ◽  
Charlotte S. Van Dorn ◽  
Jonathan N. Johnson
Keyword(s):  
Respiratory Failure ◽  
Cardiac Failure ◽  
Mechanical Circulatory Support ◽  
Ventricular Assist Devices ◽  
Circulatory Support ◽  
Systemic Diseases ◽  
Ventricular Assist ◽  
Risk Of Bleeding ◽  
Myocardial Failure ◽  
Near Future

Mechanical circulatory support (MCS) has become a critical tool for managing children with impending respiratory and cardiac failure. Although extracorporeal membrane oxygenation was classically the only form of support available for children, ventricular assist devices (VADs) are increasingly used in children. Common indications for MCS include an inability to oxygenate or ventilate that progresses to respiratory failure and cardiac failure secondary to anatomic abnormalities or primary myocardial failure. The most common contraindication for MCS is death in the near future. This contraindication may include patients with other fatal systemic diseases, patients at high risk of bleeding, or extreme prematurity. Important recent advances in VAD technology include the introduction of the Berlin Heart EXCOR device, as well as the successful use of devices for adults (e.g., the Heartmate and Heartware VADs) in larger children. Although outcomes of VAD support in children have been promising, further studies of smaller and clinically more complex children are required.

Predictive Models in Cybercrime Investigation

2010 ◽  
Vol 2 (3) ◽  
pp. 1-12 ◽  
Author(s):  
A. S. N. Murthy ◽  
Vishnuprasad Nagadevara ◽  
Rahul De'
Keyword(s):  
Data Mining ◽  
Law Enforcement ◽  
Success Rate ◽  
Predictive Models ◽  
Law Enforcement Agencies ◽  
Limited Success ◽  
Investigation Methods ◽  
The World ◽  
Policy Measures ◽  
Using Data

With increased access to computers across the world, cybercrime is becoming a major challenge to law enforcement agencies. Cybercrime investigation in India is in its infancy and there has been limited success in prosecuting the offenders; therefore, a need to understand and strengthen the existing investigation methods and systems for controlling cybercrimes is greatly needed. This study identifies important factors that will enable law enforcement agencies to reach the first step in effective prosecution, namely charge-sheeting of the cybercrime cases. Data on 300 cybercrime cases covering a number of demographic, technical and other variables related to cybercrime was analyzed using data mining techniques to identify and prioritize various factors leading to filing of the charge-sheet. These factors and the respective priority rankings are used to suggest various policy measures for improving the success rate of prosecution of cybercrimes.

scholarly journals Development of predictive models using Data Mining techniques to detect borer infestation (Diatraea saccharalis) in sugarcane culture

2017 ◽  
Author(s):  
Nádia Vieira Ribeiro ◽  
Luiz Henrique Antunes Rodrigues ◽  
Monique Pires Gravina de Oliveira ◽  
FELIPE FERREIRA BOCCA
Keyword(s):  
Data Mining ◽  
Predictive Models ◽  
Data Mining Techniques ◽  
Diatraea Saccharalis ◽  
Using Data
Sign in / Sign up

Export Citation Format

Share Document