Nowadays strawberry is one of the most popular fruits all over the world. Strawberry is so much nutrient fruits also. Strawberry fruits consist of several nutrients such as Potassium, Carbohydrate, Protein, Calcium, Magnesium, Iron and high content of Vitamin C, etc. [1]. For this reason, peoples are showing so much interest to take strawberry fruits in their routine life and the demand for strawberries is increasing day by day. In the year 2006, the total world strawberry production was 58, 41,237 tons but in the year 2014, it was hugely increased 81, 14,373 tons [1]. So, farmers are also thinking to increase the production of strawberries.

Due to the high demand for strawberries, greenhouse strawberry cultivation is growing up rapidly. For greenhouse cultivation, nowadays farmers are using different types of method. Some method is used for growing plants with soil by using mineral nutrient solutions in a water solvent. Nitrogen, Potassium, and Phosphorus (NPK) are essential for basic plant survival. Nitrogen is for cell growth, phosphorus for the roots, flowers, and buds, and potassium is for photosynthesis [2]. The basic nutrient solutions of water are considering in its composition only nitrogen, phosphorus, potassium, calcium, magnesium and sulfur, and they are supplemented with micronutrients. Greenhouse strawberry cultivation is rapidly recognizing as the most productive and efficient form of food production. In the greenhouse, farmers are using different types of nutrient solutions with water flow for strawberry productions. In our study, we try to find out the good nutrient solutions with water flow rate for getting high productions by using the support vector regression algorithm.

These study deals to analysis the nutrient solutions with water flow rate and the productions rate of strawberries. We used support vector regression algorithm to predict the average nutrient solutions with water flow rate and the average productions rate. Then we compare all nutrient solutions flow rate and the productions to find out the best water nutrition’s flow rate for getting high productions. For doing all the steps, we used strawberry greenhouse data, which has three beds and farmers used three different amounts of nutrient solutions with water flow rate for three different beds.

In statistics, support vector machines (SVMs, also support vector networks) are supervised learning models with associated learning algorithms that analyze data [3]. M.L.A.T.M. Hertog and et al, were published the predicting keeping quality of strawberries (cv. `Elsanta’) packed under modified atmospheres: an integrated model approach [4]. Carlos Mario Grijalba and et al, has published the yield of the strawberry crop under macrotunel and open field and its relation with vegetative and reproductive aspects of the plant [5]. Siyue Li and et al, has been published a paper about Water quality in relation to land use and land cover in the upper Han River Basin, China [6]. Gianulaca Caruso and et al, has been published a paper about the effects of cultural cycles and nutrient solutions on plant growth, yield and fruit quality of alpine strawberry (Fragaria vesca L) grown in hydroponics [7]. Narges Banaeian and et al, has published a paper energy and economic analysis of greenhouse strawberry production in Tehran of Iran [8]. Giorgos Mountrakis and et al, has been published a review paper about support vector machines in remote sensing [9]. Kennedy Were and et al, has been published A comparative assessment of support vector regression, artificial neural networks, and random forests for predicting and mapping soil organic carbon stocks across an Afromontane landscape [10]. Thomas Graham and et al, has been published a paper about Response of hydroponic tomato to daily applicatons of aqueous ozone via drip irrigation [11]. Huihui Yu and et al, has published prediction of the temperature in a Chinese solar greenhouse based on LSSVM optimized by improved PSO [12]. Karen Smeets and et al, has been published critical evaluation and statistical validation of a hydroponic culture system for Arabidopsis thaliana [13]. Weitang Song and et al, has been published a paper, tomato fusarium wilt and its chemical control strategies in a hydroponic system [14]. Dietmar Schwarz and et al, has been published a paper about the influence of nutrient solutions concentration and a root pathogen (Pythium aphanidermatum) on tomato root growth and morphology [15]. Erik A and et al, has been published a paper about closed soilless growing systems: A sustainable solution for Dutch greenhouse horticulture [16]. Roland De Marco and et al, has been published a paper about Hydroponic nutrient solutions using flow injection potentiometry and a cobalt-wire phosphate ion-selective electrode [17]. Ming Xie and et al, has been published a paper about Membrane-based processes for wastewater nutrient recovery: Technology, challenges, and future direction [18].

2.1 Support Vector Regression

In machine learning, support vector machines (SVMs, also support vector networks) are supervised learning models with associated learning algorithms that analyze data. Support Vector Machine can also be used as a regression method, maintaining all the main features that characterize the algorithm (maximal margin). The Support Vector Regression (SVR) uses the same principles as the SVM for classification, with only a few minor differences. First of all, because output is a real number it becomes very difficult to predict the information at hand, which has infinite possibilities. In the case of regression, a margin of tolerance (epsilon) is set in approximation to the SVM which would have already requested from the problem. But besides this fact, there is also a more complicated reason, the algorithm is more complicated therefore to be taken in consideration. However, the main idea is always the same: to minimize error, individualizing the hyperplane which maximizes the margin, keeping in mind that part of the error is tolerated. Equation of Support vector regression;

Here, y - is the output variable, x -is the number of input variable, w –is the slope of the line and b –is the error term.

In this study, we use the year 2015 September to 2016 May strawberry data, which is gained from a greenhouse strawberry farm in South Korea named Mebangsuliang. There are two types of parameters are available in the greenhouse data, first one is nutrient solutions with water flow and another one is the production rate of strawberries. Three beds are available in this greenhouse, and beds names are MBsul_A, MBsul_B, and MBsul_C. In every bed, there are six plots and in each plots, twelve plants are available. Total seventy-two plants are available in each bed. Farmers used three different types of nutrient solutions with water flow for three different beds and they got three different types of productions from three individual beds.

From, <figure 1>, 3&5 we can see the water nutrition flow rate of beds Mbsul_A, Mbsul_B & Mbsul_C and <figure 2>, 4&6 shows the production rate of beds Mbsul_A, Mbsul_ B&Mbsul_C. In the result part, we analyzed the nutrition flow rate and the production rate of Mbsul_A, Mbsul_B&Mbsul_C by using support vector regression algorithm to find out the fitting average productions of strawberries. We make a comparison of water nutrition’s follow rate among three beds and also make a comparison between the three productions rate. Finally, we find out the good nutrition flowrate for high productions.

In result and discussion part, we use the support vector regression for predicting average water nutrition’s flow rate for three beds and the average productions rate of three beds. From these results, we find out the best water nutrition’s flow rate from among three for getting high productions.

This paper focused on to find out high productions of strawberries based on water nutrient solutions flow. All results and analyze provided us acuteness between nutrient solutions with water flow and strawberry productions. All results about three different types of nutrient solutions with water flow with three different productions and this work drills to find out high productions of strawberries based on water nutrient solutions flow. We analyze absorbed data with the fitted line of support vector regression. We compare three different types of nutrient solutions with water flow among them and compare three fitted lines of support vector regression. We compare three different productions and the fitted line of support vector regression. After comparing all three beds nutrient solutions flow and productions we find out that bed Mbsul_B gives us the best strawberry productions among three beds also with an optimal water nutrient flow. The total production rate and nutrient solutions of bed Mbsul_B are 462.0009(kg) & 38396.85 (L). On the other hand, total productions of Mbsul_A and Mbsul_C are 405.6558(kg) & 407.5649(kg). Average strawberry productions and total nutrient solutions flow of two beds accordingly 9.433855(kg) &37536.44(L) for Mbsul_A and 9.478254(kg) & 41708(L) for Mbsul_C. So, the production of Mbsul_B is far better than Mbsul_A& Mbsul_C also the nutrient solutions flow of Mbsul_B is an optimum nutrient solution. Mbsul_B nutrient solutions flow is close to nutrient solutions of bed Mbsul_A and less than Mbsul_C water nutrient solutions flow. From <Figure 18>, we can know about the three beds water nutrient solutions flow. Average nutrient solutions flow also shows the difference between three beds, average nutrient solutions of Mbsul_A = 872.94(L), Mbsul_B= 892.95(L), Mbsul_C =969.96(L). After analysing all results, it also shows that the high amount of nutrient solutions flows not able to give high productions. For getting high productions, we need an optimum level of water nutrient solutions flow.