search

jlivsey / UB-sping24-time-series

3 stars 1 forks source link

Satyavathi Gunturi #4

Open vgunturi opened 5 months ago

vgunturi commented 5 months ago

https://www.kaggle.com/code/satyavathig/time-series-forecasting-uninsured-claims

  1. Data was checked for outliers using a Box plot, histogram, and IQR
  2. Outliers were dropped from the data
  3. Series was checked for stationarity using the Augmented Dickey-Fuller Test; a. p-value was less than 0.05 and |critical value| (3.43) < |absolute of test-statistic| (5.891) implies time series is stationary
  4. ACF and PACF were plotted - p, d, q, and seasonal counterparts were observed from the plots - but more clarity is needed on reading the plots (Approximate values are taken as of now)
  5. Seasonal ARIMAX was used to forecast the data (so that seasonality and economic factors are considered) The forecasted values that I have obtained significantly differ from those of my peers - Not sure if the removal of outliers is the cause
vgunturi commented 4 months ago

HW02: https://www.kaggle.com/satyavathig/forecasting-with-outlier-treatment-on-covid-data/edit

Forecasted value for 8th Feb: 203343.69 Outlier Treatment used: Box-Cox Transformation

vgunturi commented 4 months ago

HW03: https://www.kaggle.com/satyavathig/time-series-inflation/edit

Dataset Chosen: https://fred.stlouisfed.org/series/EXPINF10YR (10-year inflation data)

vgunturi commented 4 months ago

HW01: Posted on BrightSpace https://www.kaggle.com/code/satyavathig/hw-01

vgunturi commented 4 months ago

HW02_SatyavathiGunturi_TimeSeries.pdf

vgunturi commented 3 months ago

HW03_TimeSeriesAnalysis.pdf

vgunturi commented 1 month ago

https://drive.google.com/drive/folders/1FjmtPgfswX7-sPGJ1rug7M0L_bCsyR3J?usp=drive_link