機器學習是一種以自動且智慧的方式將資料轉化成有用的資訊和知識的技術。它是從大型資料庫中發現隱含的、先前 未知的以及有用的知識。機器學習強調的是建置一個完整的體系或演算機制來幫助電腦或系統自我學習，進而達到預測效果不斷提升的過程。已廣泛的應用在當前許多問題的解決上。
研究的程序是在一個區域內，紀錄26個地點所量到的無線網路基地台之訊號，而每個地點都以座標的形式表達，每個點各進行30次的測量。我們利用不同的機器學習方法來對這些已知的數據進行訓練及預測，使用的方法有:支撐向量機、k-最近鄰演算法、決策樹及隨機森林四種，利用這四種方法來從已知位置預測訊號，以及從已知訊號來預測位置，並探討不同的機器學習方法所預測的結果。

Machine learning is a technology that transforms the data to useful information and knowledge in an automated and intelligent way. It finds out the implied, unknown and useful knowledge form large database. The main purpose of the machine learning is to help computers and systems learn by themselves on a complete system or calculation rules.
Furthermore, machine learning is applied to solve many problems for improving the effect of prediction in recent years.
The procedure of this study is to record the signal of the wireless network base station from the twenty-six locations, which is expressed in coordinates and measured thirty times each within an area. We train and forecast these known data by different machine learning such as support vector machines, K-nearest-neighbor, decision tree and random decision forests. We use the four ways to predict signal by known locations or predict locations by known signal. Moreover, we investigate the forecast results of the different machine learning.

[1] Chieh-Chen Wu, "運用機器學習演算法對脂肪肝預測研究" , 臺北醫學大學醫學資訊研究所碩士論文, 2014.
[2] Li-Yuan Chiu," 應用機器學習方法預測核糖核酸與蛋白質結合位置" , 臺灣大學工程科學及海洋工程學研究所碩士論文, 2010.
[3] Wen-Cheng Chang, " 整合集群技術與機器學習技術建構電腦產品銷售預測模式" , 清雲科技大學企業管理系暨經營管理研究所碩士論文, 2012.
[4] 秦晉偉, "在無線區域網路中使用機器學習技術預測信號強度之研究" ,元智大學通訊工程學系碩士論文, 2014.
[5] J.C. Burges, "A Tutorial on Support Vector Machines for Pattern Recognition, " New York: Macmillan, 1968.
[6] An Chunming, Li Zongsen, "Studies on KNN query of moving objects for location management in spatial database," 2013 International Conference on Mechatronic Sciences, Electric Engineering and Computer (MEC) Dec 20-22, 2013
[7] G.B. Gharehpetian, "Application of core vector machines for on-line voltage security assessment using a decisiontree-based feature selection algorithm," IET Gener. Transm. Distrib.,Vol. 3, Iss. 8, pp. 701– 712,2009.
[8] Li Jun, Zhang Shunyi, Lu Yanqing, Zhang Zailong, "Internet Traffic Classification Using Machine Learning," Communications and Networking in China, 2007.
[9] J.C. Burges, "A Tutorial on Support Vector Machines for Pattern Recognition, " New York: Macmillan, 1968.
[10] 林宗勳, "Support Vector Machines 簡介"
[11] 黃子桓, " Support Vector Regression"
[12] Ibrahim Mesecan, Ihsan Ömür Bucak,"Searching the Effects of Image Scaling for Underground Object Detection Using K-Means and KNN," Modelling Symposium (EMS), 2014
[13] Liliya Demidova, Yulia Sokolova,"A novel SVM-KNN technique for classification," Embedded Computing (MECO), 2017 6th Mediterranean Conference on
[14] JiaCheng Ni, Fei Qiao, Qi Di Wu ,"A memetic PSO based KNN regression method for cycle time prediction in a wafer fab," IEEE Conference Publications, Intelligent Control and Automation (WCICA), 2012
[15] 莊斯婷, "整合決策樹與統計分析超音波抽取注入酒精在子宮模異位症療效" ,淡江大學資訊工程學系碩士論文, 2010.
[16] 林焜詳, "支撐向量機與隨機森林應用於颱風時雨量預報之比較" , 國立成功大學水利及海洋工程學系碩士論文,2016.
[17] 卓達瑋, "隨機森林分類方法於基因組顯著性檢定上之應用" ,國立政治大學統計學系碩士論文,(2010).