Abstract
According to the reference, big data is the kind of data with a great amount of capacity that cannot be analysed with the traditional database system. And the process of analysing big data helps in discovering significant hidden patterns and other regularities. Thus, machine learning algorithms are introduced to replace conventional methods in big data analysis. In this essay, the support vector machine (SVM) algorithm is applied to analyse the salary classification dataset by calculating the impact of all features in annual wage and doing the prediction based on these results. Sense for companies looking for new employees, a reasonable wage can be provided referring to their personal conditions. While for jobhunters, this prediction process can help estimate is the salary provided is worth it or not. Compared to other machine learning methods, the accuracy of SVM can be over 80% or even 90% in classification missions with both low and high-dimensional data. Additionally, varieties of kernel functions can be qualified for specific tasks. However, the running efficiency would decrease as the amount of data for training increases. And the data needs a stricter normalization step to ensure that the result would not be affected by the errors in the dataset, which could be costly. The prediction results can also be harder to interpret in the end.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have