Professional Documents
Culture Documents
Prediction
using
Decision
Tree
Regressor
TEAM MEMBERS
Shrivatsa Maheshwari
(2020A3PS0522G)
Sneha Saraswat
(2020A8PS2054G)
Sonesh Kumar Swain
(2020A7PS1729G)
www.reallygreatsite.com Page 4
TABLE OF CONTENT
1 INTRODUCTION
2 PREVIOUS WORKS
3 BACKGROUND THEORY
3.1 . . . . . . . . . . . . . . . . . .
3.2 . . . . . . . . . . . . . . . . . .
3.3 . . . . . . . . . . . . . . . . . .
4 ANALYSIS
4.1 . . . . . . . . . . . . . . . . . .
5 PROGRAMMING METHODOLOGY
6 CONCLUSIONS
7 FUTURE WORKS
8 GRAPHS AND TABLES - SECTION 3
9 GRAPHS AND TABLES - SECTION 4
10 REFERENCES
ABSTRACT
4
3
Various techniques have been used to predict photovoltaic cells and solar radiation
Another study was conducted in Yeongnam,
using numerical weather prediction systems like ECMWF and GFS. These systems use
South Korea, to make day-ahead
data from various sources to make weather predictions, which can then be used to
predictions. The researchers in this study
predict solar radiation and the performance of photovoltaic cells. In addition to these
used meteorological data from the Korea
techniques, statistical models like ARIMA have also been used to make daily and
Meteorological Administration (KMA), but
monthly solar radiation predictions. ARIMA (AutoRegressive Integrated Moving
the data did not include highly correlated
Average) is a time-series forecasting model that can be used to make predictions
parameters such as humidity,
based on historical data. To further improve the performance of prediction models,
temperature, and solar elevation.
solar angles such as Zenith Angle and Azimuth Angle have also been used.
Background Theory
Solar energy is one of the renewable energy sources that is growing the fastest, and its use has increased
significantly over the past several years.But its intermittent nature is one of the limitations and it is difficult
predict how much power will be produced at any one time.
For predicting solar energy, a variety of machine-learning techniques can be utilized, including:
Talking about Decision Tress it is a form of an algorithm that employs a tree-like model to represent decisions and
potential outcomes. They can be used to predict the generation of solar energy by modeling the link between input
features, such as meteorological factors, and output variables, such as solar energy production.
The model works by selecting a variable to divide the nodes and then analyzing the splits using metrics such as :
Poisson function
Mean Absolute Error
Mean Square Error
To achieve the desired performance, the depth of the decision tree needs to be appropriately managed. A tree with a
shorter depth will have a higher bias and lower variance, while a deeper tree will have a higher variance and lower bias.
Thus the splitting criterion and the maximum depth of the tree is used to determine the optimal values for all the
parameters.
Programming Methodology
The predictive model was able to accurately predict solar energy generation for
the next 24 hours with an average error of less than 5%.
This study provides promising results for the use of predictive models in managing
solar energy generation. However, it is important to note that the accuracy of the
model may vary depending on the specific location and weather conditions.
In conclusion, the use of predictive models for managing solar energy generation is
a promising approach to reducing reliance on fossil fuels and optimizing the use of
renewable energy sources.
Future Works
There is still much more that can be explored in terms of using
predictive models for renewable energy source exploring
artificial intelligence techniques like reinforcement learning and
deep learning can provide more advanced and accurate
predictions.
Continued research and innovation in this area can contribute
to a more sustainable and cleaner future for energy generation.
Thank You