演算法和模組參考概觀Algorithm & module reference overview

此參考內容會提供有關每個機器學習演算法和模組可供使用 Azure Machine Learning 服務的視覺化介面 (預覽) 中的技術背景。This reference content provides the technical background on each of the machine learning algorithms and modules available in the visual interface (preview) of Azure Machine Learning service.

每個模組都代表一組可獨立執行,並執行機器學習服務工作,提供必要的輸入程式碼。Each module represents a set of code that can run independently and perform a machine learning task, given the required inputs. 模組可能包含特定的演算法,或執行在機器學習服務,例如遺漏值取代或統計分析中很重要的工作。A module might contain a particular algorithm, or perform a task that is important in machine learning, such as missing value replacement, or statistical analysis.

提示

中的視覺介面的任何實驗,您可以取得特定模組的相關資訊。In any experiment in the visual interface, you can get information about a specific module. 選取的模組,然後選取更多協助連結快速說明窗格。Select the module, then select the more help link in the Quick Help pane.

模組Modules

模組會依功能:Modules are organized by functionality:

功能Functionality 描述Description 模組Module
資料格式轉換Data format conversions 在機器學習中使用的各種檔案格式的資料轉換Convert data among various file formats used in machine learning, 轉換為 CSVConvert to CSV
資料輸入和輸出Data input and output 將資料從雲端來源移到您的實驗。Move data from cloud sources into your experiment. Azure 儲存體、 SQL database,或 Hive,撰寫您的結果或中繼資料,同時執行實驗,或使用雲端儲存體來實驗之間交換資料。Write your results or intermediate data to Azure Storage, a SQL database, or Hive, while running an experiment, or use cloud storage to exchange data between experiments. 匯入資料Import Data
匯出資料Export Data
手動輸入資料Enter Data Manually
資料轉換Data transformation 是唯一的機器學習服務,例如正規化或分類收納資料、 特徵選取和維度縮減的資料作業。Operations on data that are unique to machine learning, such as normalizing or binning data, feature selection, and dimensionality reduction. 選取資料集中的資料行Select Columns in Dataset
編輯中繼資料Edit Metadata
清除遺漏的資料Clean Missing Data
加入資料行Add Columns
Add RowsAdd Rows
移除重複的資料列Remove Duplicate Rows
將資料分割Split Data
將資料標準化,Normalize Data
資料分割和取樣Partition and Sample
Python 模組Python module 撰寫程式碼,並將它內嵌至整合您的實驗中的 Python 模組中。Write code and embed it in a module to integrate Python with your experiment. 執行 Python 指令碼Execute Python Script
建立 Python 模型Create Python Model
機器學習服務演算法:Machine learning algorithms:
分類Classification 預測的類別。Predict a class. 選擇從 (雙類別) 的二進位或多級的演算法。Choose from binary (two-class) or multiclass algorithms. 多元決策樹系Multiclass Decision Forest
多級羅吉斯迴歸Multiclass Logistic Regression
多級類神經網路Multiclass Neural Network
二級羅吉斯迴歸Two-Class Logistic Regression
二級平均感知器Two-Class Averaged Perceptron
二級 促進 決策 樹狀結構Two-Class Boosted Decision Tree
二級決策樹系Two-Class Decision Forest
二級類神經網路Two-Class Neural Network
兩個‑類別 支援 向量 機器Two‑Class Support Vector Machine
叢集Clustering 群組在一起的資料。Group data together. K-means 群集K-Means Clustering
迴歸Regression 預測的值。Predict a value. 線性迴歸Linear Regression
類神經網路迴歸Neural Network Regression
決策樹系迴歸Decision Forest Regression
促進 決策 樹狀目錄 迴歸Boosted Decision Tree Regression
建置及評估模型:Build and evaluate models:
定型Train 透過演算法執行資料。Run data through the algorithm. 定型模型Train Model
定型群集模型Train Clustering Model
評估模型Evaluate Model 測量已訓練模型的精確度。Measure the accuracy of the trained model. 評估模型Evaluate Model
分數Score 從剛定型的模型取得預測。Get predictions from the model you've just trained. 套用轉換Apply Transformation
指派 資料 到 叢集Assign Data to Clusters
評分模型Score Model

錯誤訊息Error messages

深入了解錯誤訊息和例外狀況代碼您可能會遇到 Azure Machine Learning 服務的視覺化介面中使用模組。Learn about the error messages and exception codes you might encounter using modules in the visual interface of Azure Machine Learning service.