Quantasoil

QuantaSoil is a project dedicated to reducing emissions from crop production and fertilization. Issue: Excessive use of fertilizers and planting the wrong crops not only increases carbon dioxide and methane emissions into the atmosphere, but also leads to soil degradation. All of this subsequently leads to global climate change. This problem is particularly relevant for Kazakhstan, as more than 65% of the country's total exports are crops. Solution: Proper fertilization and plant selection can improve the soil's ability to absorb and store carbon, which directly reduces the amount of carbon dioxide in the atmosphere. This helps in the fight against climate change by reducing the effect of greenhouse gases. A project website where the user(farmer) is given two Machine Learning algorithms (Random Forest) to choose from. The task of the first algorithm is to use the user's data on Nitrogen, Phosphorus, and Potassium content in the soil, as well as Temperature, Humidity, soil PH, and precipitation to determine the most suitable crop to grow so that these parameters are ideal for healthy growth of the crop and the growth process of this particular crop produces the least amount of harmful emissions. The task of the second algorithm is to determine the most suitable fertilizer for growing the crop based on the user's data on Temperature, Humidity, Earth Type, Crop Type, Nitrogen, Phosphorus, and Potassium content, so that these parameters are ideal for healthy growth of the crop fertilized with this fertilizer and this fertilizer produces the least amount of harmful emissions. Both algorithms were trained on open source Datasets using self-written code using pandas, numpy, and scikit-learning libraries to train the Dataset-based Algorithms. To automate and simplify the results of the algorithms, an OpenAI API, GPT-4, was used to act as an interpreter for the Algorithms in the backend on Flask. GPT-4 interprets the output from two trained Algorithms