Dr. M.Iqbal Jeelani
Machine learning (ML) algorithms have emerged as promising alternative and complimentary tools to the commonly used modeling approaches in agriculture and allied sciences. ML algorithms have gained popularity in crop production, yield prediction and forest management research now days. Machine learning is an application of artificial intelligence that enables a system to learn from examples and experience without explicit programming. Machine learning comprises of a category of algorithms that allows software applications to become more accurate in predicting outcomes from systems of interest in research .The basic premise of ML is to build algorithms that can receive input data and use statistical analysis to predict an output while updating outputs as new data become available.
Building algorithms that can take input data and utilise statistical analysis to predict an output while updating outputs as new data become available is the fundamental tenet of machine learning. Extractions of more knowledge and picking out or recognising trends from large data sets are two aspects of machine learning and are mostly employed to handle complicated issues when human expertise fails since they can be continuously refined with greater precision. The emerging concept of Machine Learning together with big data technologies and high performance computing has created new opportunities to quantify and understand data intensive processes in new age smart farming. Now a day’sMachine learning is all over the domain of agriculture throughout the entire growing and harvesting cycle, which begins with soil preparation, seeds breeding and water feed measurement? and ultimately ends up with robots to pick up the harvest determining the readiness with the help of computer vision. Machine learning can benefit the agriculture at every stage including soil management, crop management, disease detection, livestock management ,etc.
Machine learning algorithms study evaporation processes, soil moisture and temperature to understand the dynamics of ecosystems and the impingement in agriculture. Now a day’s Machine learning based applications are used for assessment of daily, weekly, or monthly evapotranspiration allowing for a more effective use of irrigation systems and prediction of daily dew point temperature, which helps identify expected weather phenomena. The state of the art machine learning algorithms have incorporated computer vision technologies to provide data for widespread multidimensional analysis of crops, weather, and economic conditions. Apart from this Machine learning plays a very important role in weed detection which is a serious concern in traditional crop production. Detection of weeds is very challenging task as it is very difficult to detect them and differentiate them from main crop. Such challenges can be overcome by the application of MLalgorithms at low costs with no environmental issues. The algorithms like Artificial Neural Networks, Support Vector Machines, , Decision Trees, Random Forests, etc which are used in crop management processes, which are still in the beginning of its journey, have already evolved into artificial intelligence systems. ML algorithms focus on the predictive accuracy of models rather depending on the data modeling without or minimal human intervention, and can give better decision making support.
Uncertainty plays a fundamental role in all machine learning. Many aspects of it crucially depend on a careful probabilistic representation of uncertainty. One way to deal with uncertaintie seffectively is to develop probabilistic ML algorithms which can provide a framework for representing and manipulating uncertainty related to data, models, and predictions. The probabilistic ML algothirms and artificial intelligence is a very dynamic area of research with wide ranging impacts beyond conventional pattern recognition problems in agricultural production. It will continue to play a central role in the development of ever more powerful ML systems for future application in agricultural system.
(Inputs by Afshan Tabasum, Mansha Gul-both Research scholars SKUAST-Jammu)
(The author is wokring as Assistant Professor SKUAST-Jammu)