杜振鑫
摘 要:本科生数据挖掘和机器学习课程内容抽象难懂,传统的高等院校教材以讲授式为主。目前已提出大量的案例式、项目式教学方法降低学习难度,但仍然存在抽象、不容易实施、缺少对算法原理剖析的缺点。本文首次提出在数据挖掘课程中采用过程生成式启发教学,将学生的生活经验原型逐步形式化为严谨的数据挖掘算法,降低了学习难度,有助于学生对算法的深刻理解,取得了更好的教学效果,在保证学科知识教学正确性与科学性之基础上,助力学生学科素养与学科能力的生成。
关键词:数据挖掘 过程生成 生活原型 启发式教学
中图分类号:TP312.1 文献标识码:A 文章编号:
Abstract: The content of undergraduate data mining and machine learning course is abstract and difficult to understand. The traditional teaching materials of colleges and universities are mainly lecture teaching. At present, a large number of case-based and project-based teaching methods have been proposed to reduce the learning difficulty. However, there are still some shortcomings, such as abstraction, not easy to implement and lack of analysis of the principle of the algorithm. In this paper, the process generation heuristic teaching is proposed for the first time in the data mining course, and the prototype of students' life experience is gradually formalized into a rigorous data mining algorithm, which reduces the learning difficulty, contributes to students' deep understanding of the algorithm and achieves better teaching results, on the basis of ensuring the correctness and scientificity of subject knowledge teaching, help students generate subject literacy and subject ability.
Key Words: Data mining; Processgeneration; Life prototype; Heuristic teaching
隨着各行各业技术的发展,这个时代的数据量已经发生跨越式的增长[1][2]。预计到 2025 年,全球天文数据采集量将达到每年 2.5×10 10 TB;仅2017 年,淘宝每天产生的数据都高达 7TB;物联网中的每台设备都会产生大量的数据[1]。 IDC 在《2020 年的“数字宇宙”》研究报告中预测,到 2020 年,全球数据量将达到 40ZB,中国在全球数据市场的份额也将由目前的 13% 上升到 22%[2]。大数据技术的出现很好地解决了大量数据的计算问题,数据挖掘和机器学习也成为一门非常重要的课程。但是数据挖掘和机器学习教学存在抽象、难于理解的问题。传统教材大多采用算法加例题的讲授模式,虽然表述严谨,却不利于教学。学生突兀的接受某个算法,只见树木不见森林,往往产生厌学情绪,不利于人才培养。覃凤萍[1]提出问题引导+案例教学法,高建瓴[3]提出问题驱动教学模式,这些方法都是通过案例进行教学,能够有效降低学生的学习难度。……