Using the TSA-LSTM two-stage model to predict cancer incidence and mortality

Cancer, the second-leading cause of mortality, kills 16% of people worldwide. Unhealthy lifestyles, smoking, alcohol abuse, obesity, and a lack of exercise have been linked to cancer incidence and mortality. However, it is hard. Cancer and lifestyle correlation analysis and cancer incidence and mortality prediction in the next several years are used to guide people’s healthy lives and target medical financial resources. Two key research areas of this paper are Data preprocessing and sample expansion design Using experimental analysis and comparison, this study chooses the best cubic spline interpolation technology on the original data from 32 entry points to 420 entry points and converts annual data into monthly data to solve the problem of insufficient correlation analysis and prediction. Factor analysis is possible because data sources indicate changing factors. TSA-LSTM Two-stage attention design a popular tool with advanced visualization functions, Tableau, simplifies this paper’s study. Tableau’s testing findings indicate it cannot analyze and predict this paper’s time series data. LSTM is utilized by the TSA-LSTM optimization model. By commencing with input feature attention, this model attention technique guarantees that the model encoder converges to a subset of input sequence features during the prediction of output sequence features. As a result, the model’s natural learning trend and prediction quality are enhanced. The second step, time performance attention, maintains We can choose network features and improve forecasts based on real-time performance. Validating the data source with factor correlation analysis and trend prediction using the TSA-LSTM model Most cancers have overlapping risk factors, and excessive drinking, lack of exercise, and obesity can cause breast, colorectal, and colon cancer. A poor lifestyle directly promotes lung, laryngeal, and oral cancers, according to visual tests. Cancer incidence is expected to climb 18–21% between 2020 and 2025, according to 2021. Long-term projection accuracy is 98.96 percent, and smoking and obesity may be the main cancer causes.