Short-term Traffic Flow Prediction DSSA With Mapreduce

In short-term traffic flow prediction,researchers are facing the challenge of “bigdata,large amounts of computation”.And they have to develop implement for everyalgorithm.In this paper, Traffic Predition DSSA With MapReduce is proposed toresolve not only the problem of the “big data,large amounts of computation” but alsothe problem of software reuse,both in component level and architechture level.Theachievements are below:1.An algorithm architechture is proposed to make the software able to handle allth time-series-based algorithms.2.A domain model is developed for short-term traffic flow prediction.3.A short-term traffic flow prediction is designed for software development inthis area.In conclusion, the computation model of MapReduce and Hadoop MapReducehelps to handle big data and the large scale of computation upon the big data.Whilethe algorithm architechture helps to make the software able to handle all thetime-series-based algorithms.Finally,the DSSA is designed base on the model and thealgorithm architechture to help develop software more easily.