[数据挖掘] 48 时间序列预测 7 平滑法-复杂指数平滑法 Holt's linear method和 Holt-Winter method 美铁数据_哔哩哔哩_bilibili
# holt-winters 指数平滑法直接应用于美铁客流量数据
expsmoothFit= ExponentialSmoothing(train_ts,trend='additive',seasonal='additive',seasonal_periods=12,freq='MS').fit()
fig,axes=plt.subplots(nrows=2,ncols=1,figsize=(9,8),dpi=180)
expsmoothFit.fittedvalues.plot(ax=axes[0],linewidth=2,color='plum')
expsmoothFit.forecast(len(test_ts)).plot(ax=axes[0],linewidth=2,color='plum',style='--')
residual = train_ts - expsmoothFit.fittedvalues
residual.plot(ax=axes[1],color='mediumvioletred')
residual = test_ts-exps