{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"\n",
"
\n",
" An introduction to Machine Learning with Scikit-Learn
\n",
"
\n",
" Yifei Huang\n",
"
\n",
" MAFS6010U\n",
"
\n",
" HKUST\n",
""
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Prerequisites "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"- Materials available on Piazza\n",
" \n",
"- Require a Python distribution with scientific packages (NumPy, SciPy, Scikit-Learn, Pandas)\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Outline\n",
"\n",
"* Preprocess data with Pandas and Numpy in Python\n",
"* Scikit-Learn and Machine Learning in Python\n",
"* Various Classification Models\n",
"* Model evaluation and selection\n",
"* Summary"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Load some packages"
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Collecting mpl_finance\n",
"Requirement already satisfied: matplotlib in /home/yhuangcc/anaconda3/envs/mafs6010u/lib/python3.7/site-packages (from mpl_finance) (3.0.2)\n",
"Requirement already satisfied: numpy>=1.10.0 in /home/yhuangcc/anaconda3/envs/mafs6010u/lib/python3.7/site-packages (from matplotlib->mpl_finance) (1.15.4)\n",
"Requirement already satisfied: cycler>=0.10 in /home/yhuangcc/anaconda3/envs/mafs6010u/lib/python3.7/site-packages (from matplotlib->mpl_finance) (0.10.0)\n",
"Requirement already satisfied: kiwisolver>=1.0.1 in /home/yhuangcc/anaconda3/envs/mafs6010u/lib/python3.7/site-packages (from matplotlib->mpl_finance) (1.0.1)\n",
"Requirement already satisfied: pyparsing!=2.0.4,!=2.1.2,!=2.1.6,>=2.0.1 in /home/yhuangcc/anaconda3/envs/mafs6010u/lib/python3.7/site-packages (from matplotlib->mpl_finance) (2.3.0)\n",
"Requirement already satisfied: python-dateutil>=2.1 in /home/yhuangcc/anaconda3/envs/mafs6010u/lib/python3.7/site-packages (from matplotlib->mpl_finance) (2.7.5)\n",
"Requirement already satisfied: six in /home/yhuangcc/anaconda3/envs/mafs6010u/lib/python3.7/site-packages (from cycler>=0.10->matplotlib->mpl_finance) (1.12.0)\n",
"Requirement already satisfied: setuptools in /home/yhuangcc/anaconda3/envs/mafs6010u/lib/python3.7/site-packages (from kiwisolver>=1.0.1->matplotlib->mpl_finance) (40.6.3)\n",
"Installing collected packages: mpl-finance\n",
"Successfully installed mpl-finance-0.10.0\n"
]
}
],
"source": [
"import sklearn\n",
"import numpy as np\n",
"import pandas as pd\n",
"try:\n",
" from mpl_finance import candlestick_ohlc\n",
"except ImportError:\n",
" ! pip install mpl_finance\n",
" from mpl_finance import candlestick_ohlc\n",
"import matplotlib.pyplot as plt\n",
"import matplotlib.dates as mdates\n",
"plt.style.use('seaborn')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Loading external data with pandas"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"data=pd.read_csv('./okex_future_BTC_USD_this_week_1H.csv',parse_dates=['candle_begin_time'])"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n",
"
\n",
" \n",
" \n",
" | \n",
" candle_begin_time | \n",
" open | \n",
" high | \n",
" low | \n",
" close | \n",
" volume | \n",
"
\n",
" \n",
" \n",
" \n",
" | 0 | \n",
" 2018-05-15 00:00:00 | \n",
" 8674.33 | \n",
" 8761.76 | \n",
" 8663.17 | \n",
" 8696.66 | \n",
" 172152.0 | \n",
"
\n",
" \n",
" | 1 | \n",
" 2018-05-15 01:00:00 | \n",
" 8694.66 | \n",
" 8728.17 | \n",
" 8662.17 | \n",
" 8701.85 | \n",
" 122668.0 | \n",
"
\n",
" \n",
" | 2 | \n",
" 2018-05-15 02:00:00 | \n",
" 8699.98 | \n",
" 8741.04 | \n",
" 8642.36 | \n",
" 8673.34 | \n",
" 181228.0 | \n",
"
\n",
" \n",
" | 3 | \n",
" 2018-05-15 03:00:00 | \n",
" 8673.19 | \n",
" 8717.70 | \n",
" 8638.95 | \n",
" 8707.10 | \n",
" 193166.0 | \n",
"
\n",
" \n",
" | 4 | \n",
" 2018-05-15 04:00:00 | \n",
" 8710.92 | \n",
" 8737.36 | \n",
" 8684.20 | \n",
" 8699.16 | \n",
" 155912.0 | \n",
"
\n",
" \n",
"
\n",
"
"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"data['timestamp']=data.candle_begin_time.apply(mdates.date2num)\n",
"fig,ax=plt.subplots(1,1,figsize=(40,10))\n",
"plt.title('OKEX-FUTURE_BTC-USD_1H',fontsize=30)\n",
"candlestick_ohlc(ax,data[['timestamp','open','high','low','close']].values,width=.6,colorup='g',alpha=1)\n",
"ax.xaxis_date()\n",
"ax.xaxis.set_major_formatter(mdates.DateFormatter('%y-%m-%d'))\n",
"ax.set_ylabel('Price',fontdict={'size':30})\n",
"ax.set_xlabel('Date',fontdict={'size':30})"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### What we want to do is to predict the next price down or up based on last price"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"data['down_or_up']=np.sign(data.close.diff().shift(-1))\n",
"data=data.dropna()"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n",
"
\n",
" \n",
" \n",
" | \n",
" candle_begin_time | \n",
" open | \n",
" high | \n",
" low | \n",
" close | \n",
" volume | \n",
" timestamp | \n",
" down_or_up | \n",
"
\n",
" \n",
" \n",
" \n",
" | 0 | \n",
" 2018-05-15 00:00:00 | \n",
" 8674.33 | \n",
" 8761.76 | \n",
" 8663.17 | \n",
" 8696.66 | \n",
" 172152.0 | \n",
" 736829.000000 | \n",
" 1.0 | \n",
"
\n",
" \n",
" | 1 | \n",
" 2018-05-15 01:00:00 | \n",
" 8694.66 | \n",
" 8728.17 | \n",
" 8662.17 | \n",
" 8701.85 | \n",
" 122668.0 | \n",
" 736829.041667 | \n",
" -1.0 | \n",
"
\n",
" \n",
" | 2 | \n",
" 2018-05-15 02:00:00 | \n",
" 8699.98 | \n",
" 8741.04 | \n",
" 8642.36 | \n",
" 8673.34 | \n",
" 181228.0 | \n",
" 736829.083333 | \n",
" 1.0 | \n",
"
\n",
" \n",
" | 3 | \n",
" 2018-05-15 03:00:00 | \n",
" 8673.19 | \n",
" 8717.70 | \n",
" 8638.95 | \n",
" 8707.10 | \n",
" 193166.0 | \n",
" 736829.125000 | \n",
" -1.0 | \n",
"
\n",
" \n",
" | 4 | \n",
" 2018-05-15 04:00:00 | \n",
" 8710.92 | \n",
" 8737.36 | \n",
" 8684.20 | \n",
" 8699.16 | \n",
" 155912.0 | \n",
" 736829.166667 | \n",
" 1.0 | \n",
"
\n",
" \n",
"
\n",
"
\n",
"\n",
"
\n",
" \n",
" \n",
" | \n",
" candle_begin_time | \n",
" open | \n",
" high | \n",
" low | \n",
" close | \n",
" volume | \n",
" timestamp | \n",
" down_or_up | \n",
"
\n",
" \n",
" \n",
" \n",
" | 6618 | \n",
" 2019-02-14 18:00:00 | \n",
" 3574.04 | \n",
" 3576.45 | \n",
" 3573.22 | \n",
" 3575.91 | \n",
" 210.194723 | \n",
" 737104.750000 | \n",
" -1.0 | \n",
"
\n",
" \n",
" | 6619 | \n",
" 2019-02-14 19:00:00 | \n",
" 3576.18 | \n",
" 3577.88 | \n",
" 3571.17 | \n",
" 3574.89 | \n",
" 164.462093 | \n",
" 737104.791667 | \n",
" -1.0 | \n",
"
\n",
" \n",
" | 6620 | \n",
" 2019-02-14 20:00:00 | \n",
" 3576.06 | \n",
" 3576.06 | \n",
" 3563.81 | \n",
" 3566.38 | \n",
" 178.780002 | \n",
" 737104.833333 | \n",
" 1.0 | \n",
"
\n",
" \n",
" | 6621 | \n",
" 2019-02-14 21:00:00 | \n",
" 3565.87 | \n",
" 3575.20 | \n",
" 3556.68 | \n",
" 3572.00 | \n",
" 218.644452 | \n",
" 737104.875000 | \n",
" -1.0 | \n",
"
\n",
" \n",
" | 6622 | \n",
" 2019-02-14 22:00:00 | \n",
" 3570.56 | \n",
" 3572.85 | \n",
" 3564.21 | \n",
" 3566.40 | \n",
" 192.240609 | \n",
" 737104.916667 | \n",
" -1.0 | \n",
"
\n",
" \n",
"
\n",
"