An Extensive Dataset for the Heart Disease Classification System
Description
Finding a good data source is the first step toward creating a database. Cardiovascular illnesses (CVDs) are the major cause of death worldwide. CVDs include coronary heart disease, cerebrovascular disease, rheumatic heart disease, and other heart and blood vessel problems. According to the World Health Organization, 17.9 million people die each year. Heart attacks and strokes account for more than four out of every five CVD deaths, with one-third of these deaths occurring before the age of 70 A comprehensive database for factors that contribute to a heart attack has been constructed , The main purpose here is to collect characteristics of Heart Attack or factors that contribute to it. As a result, a form is created to accomplish this. Microsoft Excel was used to create this form. Figure 1 depicts the form which It has nine fields, where eight fields for input fields and one field for output field. Age, gender, heart rate, systolic BP, diastolic BP, blood sugar, CK-MB, and Test-Troponin are representing the input fields, while the output field pertains to the presence of heart attack, which is divided into two categories (negative and positive).negative refers to the absence of a heart attack, while positive refers to the presence of a heart attack.Table 1 show the detailed information and max and min of values attributes for 1319 cases in the whole database.To confirm the validity of this data, we looked at the patient files in the hospital archive and compared them with the data stored in the laboratories system. On the other hand, we interviewed the patients and specialized doctors. Table 2 is a sample for 1320 cases, which shows 44 cases and the factors that lead to a heart attack in the whole database,After collecting this data, we checked the data if it has null values (invalid values) or if there was an error during data collection. The value is null if it is unknown. Null values necessitate special treatment. This value is used to indicate that the target isn’t a valid data element. When trying to retrieve data that isn't present, you can come across the keyword null in Processing. If you try to do arithmetic operations on a numeric column with one or more null values, the outcome will be null. An example of a null values processing is shown in Figure 2.The data used in this investigation were scaled between 0 and 1 to guarantee that all inputs and outputs received equal attention and to eliminate their dimensionality. Prior to the use of AI models, data normalization has two major advantages. The first is to avoid overshadowing qualities in smaller numeric ranges by employing attributes in larger numeric ranges. The second goal is to avoid any numerical problems throughout the process.After completion of the normalization process, we split the data set into two parts - training and test sets. In the test, we have utilized1060 for train 259 for testing Using the input and output variables, modeling was implemented.