Shulou(Shulou.com)06/01 Report--

Today, I will talk to you about the application practice of Smartbi data mining tools, which may not be well understood by many people. in order to make you understand better, the editor has summarized the following content for you. I hope you can get something according to this article.

For a successful data mining project, in addition to mature and easy-to-use products, it is more important to have a professional consulting and implementation team. What kind of software is used in data mining? Sematic software has rich practical experience in data mining and has many successful cases in different industries.

Case 1: clinical diagnosis of novel coronavirus infection

During the outbreak, novel coronavirus mainly took the positive nucleic acid test as the diagnostic standard. However, nucleic acid detection is inefficient and cannot guarantee 100% accuracy.

Novel coronavirus infection can also be identified by doctors' clinical diagnosis, but doctors are required to have rich clinical diagnosis experience.

Then, how to make ordinary doctors have rich experience in a short period of time will play an important role in the rapid prevention and control of the epidemic.

At this time, the "novel coronavirus infection Clinical diagnosis system" developed by Sematic software based on Smartbi Mining came in handy.

Smartbi Mining data mining system uses artificial intelligence machine learning algorithm to learn and train the clinical diagnosis experience of senior doctors, train to complete the novel coronavirus infection recognition model with high accuracy, and assist doctors to make clinical diagnosis quickly.

In the event of similar public health emergencies, it can buy time for rapid diagnosis, rapid isolation and rapid treatment of patients, and reduce the speed of virus transmission and patient mortality. The system can quickly train and complete the novel coronavirus infection diagnosis and identification model, and has mature support for the whole process of data access, data processing, model training, model evaluation, model deployment and so on. Smartbi Mining data mining system relies on data including epidemiological history survey data, patient clinical performance data and some laboratory data, and directly uses these features as the characteristic data of the training model. The loaded feature data is divided according to the proportion of 7:3, 70% as the training model data and 30% as the test verification model data. The logical regression algorithm and gradient lifting decision tree in the classification model are selected for diagnosis model training. 30% of the split data is used for model verification to verify the ability of the trained diagnostic model to predict new data.

Case 2: early warning of default risk of banks

The bank's "enterprise default risk early warning" project, based on the settlement behavior of public customers, including transaction frequency, transaction amount, counterparty and other information as important basic information, combined with customer industry, scale, operating conditions to depict customer portraits.

Use the logical regression model provided by Smartbi Mining data mining system to build customer overdue and default early warning. After the model is mature, use the CRM system to get through the customer manager notification channel, push the early warning data to the customer manager in time, and do a good job of risk management.

Case 3: accurate marketing of insurance customers

In the Smartbi Mining data mining system, we first select the customers who are ready for marketing through the data source, and then subdivide the customers based on the configured conditions.

These operations can be achieved by dragging the corresponding function icon and then simply configuring the parameters. The next step is to carry out different content communication according to different customer groups, which needs to be calculated through the model algorithm combined with the customer's historical data, and then to judge whether the customer has clicked any links in the SMS or left any information. After the data is returned, the following node will make a judgment, and then automatically track the implementation results of the marketing.

After reading the above, do you have any further understanding of the application and practice of Smartbi data mining tools? If you want to know more knowledge or related content, please follow the industry information channel, thank you for your support.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.