city-air-test

The main parameters of the electrochemical sensor SO2 are the voltage values at the working and auxiliary electrodes SO2op1 and SO2op2. In addition to the main parameters, the sensor can be affected by the temperature and concentration of the cross gas NO2. There are two files at your disposal:

• File with the results of laboratory calibration of the sensor - G2.csv.
• File with sensor readings in the field - test.csv The examinee is asked to propose a model describing the concentration of SO2 in the atmosphere taking into account the influence of external factors on the sensor and build a model forecast based on data collected in the field. Task by points:

1. Read the files G2.csv and test.csv.
2. Propose a candidate model (with the exception of multiparameter linear regression) describing the dependence of the gas concentration SO2 on the values of SO2op1, SO2op2, T and cross-gas NO2.
3. Compare the forecast of the resulting model with multivariable linear regression. The choice of comparison metrics is at the discretion of the examinee.
4. Construct graphs of model forecasts versus time on training and test samples.
5. Interpret the results obtained.
6. Draw conclusions.

---

• Additional questions:
  • Evaluate the errors of the proposed model and linear regression.
  • How does the proposed model differ from linear regression?
  • What can be said about the training and test samples?