توضیحات
ABSTRACT
Orange is a machine learning and data mining suite for data analysis through Python scripting and visual programming. Here we report on the scripting part, which features interactive data analysis and component-based assembly of data mining procedures. In the selection and design of components, we focus on the flexibility of their reuse: our principal intention is to let the user write simple and clear scripts in Python, which build upon C++ implementations of computationallyintensive tasks. Orange is intended both for experienced users and programmers, as well as for students of data mining.
INTRODUCTION
Scripting languages have recently risen in popularity in all fields of computer science. Within the context of explorative data analysis, they offer advantages like interactivity and fast prototyping by gluing together existing components or adapting them for new tasks. Python is a scripting language with clear and simple syntax, which also made it popular in education. Its relatively slow execution can be circumvented by using libraries that implement the computationally intensive tasks in lowlevel languages. Python offers a huge number of extension libraries. Many are related to machine learning, including several general packages like scikit-learn (Pedregosa et al., 2011), PyBrain (Schaul et al., 2010) and mlpy (Albanese et al., 2012). Orange was conceived in late 1990s and is among the oldest of such tools. It focuses on simplicity, interactivity through scripting, and component-based design. Orange library is a hierarchically-organized toolbox of data mining components. The low-level procedures at the bottom of the hierarchy, like data filtering, probability assessment and feature scoring, are assembled into higher-level algorithms, such as classification tree learning. This allows developers to easily add new functionality at any level and fuse it with the existing code. The main branches of the component hierarchy are: data management and preprocessing for data input and output, data filtering and sampling, imputation, feature manipulation (discretization, continuization, normalization, scaling and scoring), and feature selection, classification with implementations of various supervised machine learning algorithms (trees, forests, instance-based and Bayesian approaches, rule induction), borrowing from some well-known external libraries such as LIBSVM (Chang and Lin, 2011), regression including linear and lasso regression, partial least square regression, regression trees and forests, and multivariate regression splines, association for association rules and frequent itemsets mining, ensembles implemented as wrappers for bagging, boosting, forest trees, and stacking, clustering, which includes k-means and hierarchical clustering approaches, evaluation with cross-validation and other sampling-based procedures, functions for scoring the quality of prediction methods, and procedures for reliability estimation, projections with implementations of principal component analysis, multi- dimensional scaling and self-organizing maps.
Year: 2013
Publishe: University of Ljubljana
By: Usama Fayyad, Gregory Piatetsky-Shapiro, Padhraic Smyth
File Information: English Language/ 5 Page / size:67KB
Download: click
سال : 2013
ناشر : University of Ljubljana
کاری از : Usama Fayyad, Gregory Piatetsky-Shapiro, Padhraic Smyth
اطلاعات فایل : زبان انگلیسی / 5 صفحه / حجم : 67KB
لینک دانلود : روی همین لینک کلیک کنید
نقد و بررسیها
هنوز بررسیای ثبت نشده است.