The temperatures have been constantly rising over decades. The increase in the rise of temperature has led to the extinction or endangerment of many species from the earth. In this project, I will visually look at the increase in temperature globally by comparing the temperature values. The average temperature of the earth has been increased drastically. We are able to experience that many animal species are facing extinction due to global warming. However, it is still achievable by imposing certain methods by all of us toward the control of the temperature. Therefore, this project is using the weather data to demonstrate the increase in temperatures in vivid demonstrations by using data visualization techniques like visual representation of temperatures using bar graph, scatter plots, heat maps, distribution charts, etc. The data visualization techniques are very effective and help the datasets to visualize the numerical effectively. The RNN (recurrent neural networks) and ReLU (rectified linear unit) are used to predict the average rise in temperatures for the next 40–50 years. We will also use panda SQL to determine the hottest country and identify where the temperature is rising rapidly. The limitation to this project is that the LSTM cannot be trained properly due to the difference in convergence of values. The next potential steps of the project are to find the average rise in temperatures based on altitude, latitude, and carbon emissions released by them for specific areas or regions.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Temperature Prediction Using Recurrent Neural Networks and Rectified Linear Unit

  • Bandaru Shanmukha Priya,
  • Akuthota Rajashekar Reddy

摘要

The temperatures have been constantly rising over decades. The increase in the rise of temperature has led to the extinction or endangerment of many species from the earth. In this project, I will visually look at the increase in temperature globally by comparing the temperature values. The average temperature of the earth has been increased drastically. We are able to experience that many animal species are facing extinction due to global warming. However, it is still achievable by imposing certain methods by all of us toward the control of the temperature. Therefore, this project is using the weather data to demonstrate the increase in temperatures in vivid demonstrations by using data visualization techniques like visual representation of temperatures using bar graph, scatter plots, heat maps, distribution charts, etc. The data visualization techniques are very effective and help the datasets to visualize the numerical effectively. The RNN (recurrent neural networks) and ReLU (rectified linear unit) are used to predict the average rise in temperatures for the next 40–50 years. We will also use panda SQL to determine the hottest country and identify where the temperature is rising rapidly. The limitation to this project is that the LSTM cannot be trained properly due to the difference in convergence of values. The next potential steps of the project are to find the average rise in temperatures based on altitude, latitude, and carbon emissions released by them for specific areas or regions.