Perpetuation towards society in transfer technology in BMKG semarang
Posted on: 2018-09-25 22:28:10
Transfer technology data mining and machine learning to predict the weather in BMKG semarang by the computer science/informatics team of UNDIP
Weather is an environment condition, whether its hot or cold, moist or dry, high intensity rain or medium and many more. Weather prediction was one of the most challenging task to do because there is so many factor ranging from nature as well as human factor that influence the change in weather such as season, greenhouse effect, deforestation and etc. Weather prediciton plays an important role in many component in decision making regarding many subject such as agriculture, business, tourism, energy management, health whether it is humaan being or animalmand etc. The determining factor on climate consist on how to anticipate the current situation with the air changing as the atmosphere condition display was taken by ground recognition, for example from ship, plane radio sound, doppler radar,and satellite.
The centre of data and information in meteorology, climatology, and geophysics (BMKG) semarang has a purpose to give data and information service regarding weather prediction in central java. The prediction information of rainfall in central java was needed by the society as a platformto determine policies in many sector such as business, agriculture, tourism, transportation and construction.
In predicting weather at BMKG Semarang, in the early stages data and information would be assembled by each weather observation station which would then sent into the meteorology centre where the data would be assembled, checked, and then be made into a framework, map, and diagram. In BMKG Semarang, the prediction only use a few method which most of them still uses the singe parameter and analysis by the climatologist. Data mining and machine learning has developed rapidly in the last few decades to predict the weather. Predictive analysis has reach a whole new level with machine learning technique. The weather data that was used by this technique depends on its characteristic and because of that, the effective estimation was needed to share quantitatively with utilizing probability theory
As the part of the three dharma of university, the informatics/computer science perpetuation team UNDIP was doing the perpetuation that was led by Dr,E,ng. Adi wibowo, S.Si, M.Kom titled “software technology transfer data mining and machine learning for weather prediction BMKG Semarang” that was held on 27 october 2017 at BMKG station Semarang, Siliwangi street Semarang. This perpetuation strive to socialize and experiment the ysage of data mining and machine learning technique using open source software to predict the weather. In this activity k-nearest neighbour and neural network method was introduced to the participant as well as rapidminer software to be experimented. The participant for this event was data centre staff and information of BMKG Semarang that has the obligation to provide the information of weather prediction every month, here is the documentation of the event.
