Nairobi Securities Exchange All Stocks Prices 2007-2012

Nairobi Securities Exchange All Stocks Prices 2007-2012

This historical data is valuable for machine learning algorithms that need data for training and testing. It was initially compiled as part of a project to apply Artificial Neural Networks (ANN) for next day stock price prediction, based on the prices of the previous five days. Use of ANN was to contrast to prevalent methods of stock market prediction such as technical, fundamental or time series analysis.

This data was used for an initial research [1],[2] that tested an ANN for stock market prediction. This ANN model was of configuration - feedforward multi-layer perceptron (MLP) with error backpropagation, using sigmoid activation function, with network configuration 5:21:21:1. The data used in the research was for the 5-year period (2008 to 2012), with 80% of the data (4-year data) used for training and the balance 20% used for testing (last 1-year data). However, the full data that had been compiled, as presented here is the 6-year data for period 2007 to 2012.

This 2007 to 2012 data for the Nairobi Securities Exchange (NSE) was scrapped from a public website that publishes daily stock prices and archives historical data [3]. The raw data was first exported to a spreadsheet then headers, footers and other unnecessary elements were removed. This data consists of daily stock prices for over 60 different stocks and market indices over a 12-month period in each year, for a total of 6-years. Each data row has the following 13 data columns (1) Date of trade (2) Code of the stock (3) Name of the stock (4) 12-month Low price (5) 12-month High price (6) Day's Low price (7) Day's High price (8) Day's Final Price (9) Previous traded price (10) Change in price by value (11) Change in price by % (12) Volume traded (13) Adjusted price, if any. The original data also had a column on the price direction, being an arrow graphic, showing Up, Down or Unchanged. This column was discarded from this final compilation.

While the initial research only tested prediction based on adjusted final price for six stocks, the publishing of the full data now gives other researchers the opportunity to test any other of the more than 60 stocks and test any other hypothesis on this full data set. This data can also be used to reproduce and validate the initial research findings as already published.

List of data files on this dataset:
NSE_data_all_stocks_2007.csv
NSE_data_all_stocks_2008.csv
NSE_data_all_stocks_2009.csv
NSE_data_all_stocks_2010.csv
NSE_data_all_stocks_2011.csv
NSE_data_all_stocks_2012.csv

References:
[1] Wanjawa, B. W. (2014). A Neural Network Model for Predicting Stock Market Prices at the Nairobi Securities Exchange (Dissertation, University of Nairobi).
[2] Wanjawa, B. W., & Muchemi, L. (2014). ANN model to predict stock prices at stock exchange markets. arXiv preprint arXiv:1502.06434.
[3] Synergy Systems Ltd. (2020). MyStocks. Retrieved March 9, 2020, from http://live.mystocks.co.ke/

内罗毕证券交易所全股票价格数据集（2007-2012年）

该历史数据集对需用于训练与测试的机器学习算法具有重要应用价值。本数据集最初作为一项研究项目的组成部分被构建，旨在基于前五个交易日的股价，应用人工神经网络（Artificial Neural Networks, ANN）实现次日股价预测，以对比当前主流的股市预测方法，包括技术分析、基本面分析与时间序列分析。

本数据集曾被用于两项初始研究[1][2]，用以验证人工神经网络在股市预测中的应用效果。该人工神经网络模型采用带误差反向传播的前馈多层感知机（Feedforward Multi-Layer Perceptron, MLP）架构，激活函数为Sigmoid函数，网络配置为5:21:21:1。原研究使用的数据集为2008至2012年的5年数据，其中80%（即4年数据）用于模型训练，剩余20%（最后1年数据）用于模型测试。但本次发布的完整数据集涵盖了2007至2012年共计6年的历史数据。

本2007-2012年内罗毕证券交易所（Nairobi Securities Exchange, NSE）数据集爬取自一家发布每日股价并归档历史数据的公开网站[3]。原始数据首先被导出至电子表格，随后移除了表头、页脚及其他无关元素。数据集包含60余种不同股票与市场指数的每日股价信息，统计周期为每年12个月，累计覆盖6年时间。每条数据行包含以下13列数据：(1) 交易日期；(2) 股票代码；(3) 股票名称；(4) 12个月最低价；(5) 12个月最高价；(6) 当日最低价；(7) 当日最高价；(8) 当日收盘价；(9) 前一交易日成交价；(10) 股价变动额；(11) 股价变动百分比；(12) 成交量；(13) 调整后收盘价（若有）。原始数据中还包含一列以箭头图形表示的价格走势方向，分别代表上涨、下跌或持平，但该列已在最终数据集中被移除。

尽管初始研究仅针对6只股票的调整后收盘价进行了预测实验，但本次发布的完整数据集为其他研究者提供了测试超过60种股票及验证其他研究假设的可能。同时，该数据集也可用于复现并验证已发表的初始研究结果。

本数据集包含的数据文件如下：
NSE_data_all_stocks_2007.csv
NSE_data_all_stocks_2008.csv
NSE_data_all_stocks_2009.csv
NSE_data_all_stocks_2010.csv
NSE_data_all_stocks_2011.csv
NSE_data_all_stocks_2012.csv

参考文献：
[1] Wanjawa, B. W. (2014). 内罗毕证券交易所股价预测的神经网络模型[学位论文]. 内罗毕大学.
[2] Wanjawa, B. W., & Muchemi, L. (2014). 用于预测证券交易所市场股价的人工神经网络模型. arXiv预印本arXiv:1502.06434.
[3] Synergy Systems Ltd. (2020). MyStocks. 2020年3月9日检索自 http://live.mystocks.co.ke/