Parallel Patient Treatment Time Prediction Using Effective Queuing Recommendation System

As we see the increasing population and various types of emerging health related issues. These issues are difficult to predict which is a difficult job because different check-ups and a variety of tests eg., a sugar level or a blood test ,x-rays ,C.T. scan etc. All these issues and lack of patient queue management make these hospitals over crowded. Hence, most patients has to wait for long time in queue‘s, has to face complications in understanding whether which operation/check-ups they have to undergo. We are targeting to serve the patients to complete their treatment /check-ups in a very sure time and serving hospitals to schedule each treatment , operation/check-ups to avoid overcrowding of hospitals. We are going to use large realistic knowledge from various hospitals to develop patient treatment time consumption model. The realistic patient knowledge are determined by supported necessary specifications , like patient treatment starting time, end time, patient age and other useful details for each different task. We are likely to analyse and simulate totally varying waiting times for different patients by seeing their conditions and operations which are carried out by taking treatment

Accordingto literature survey after looking at various IEEE paper, we gathered some identical papers and documents. Some of the topics are discussed here: “Self-Adaptive Induction of Regression Trees”,Rau´ l Fidalgo-Merino and Marlon Nu´nez:- This Algorithm called as SAIRT used for incremental construction of binary regression trees. It mainly used to handles both symbolic and numeric attributes. The proposed algorithm can automatically adapt its internal parameters and it used to obtain model structure depending on the current dynamics of the data stream. SAIRT can monitor the usefulness of nodes and can forget examples from selected regions, storing the remaining ones in local windows associated to the leaves of the tree. On these conditions, current regression methods need a careful configuration depending on the dynamics of the problem. In this proposed algorithm obtains better results than current algorithms when dealing with data streams that involve changes with different speeds, noise levels, sampling distribution of examples, and partial or complete changes of the underlying function. “Parallel Boosted Regression Trees for net Search Ranking”,Stephen Tyree, Kilian Q.Weinberger, Kunal Agrawal:- GBT is a machine learning algorithms widely used in industry, and data analytics competitions. In this paper we propose a scalable execution plan using the parameter server architecture to facilitate the model aggregation. We also propose a work partition strategy, and a sparse-pull method, an efficient index structure to increase the processing speed. We also presented a parallel algorithm for training gradient boosted regression

Correlation based mostly ripping criterionin multi branch call tree”, bureau Salehi-Moghaddami_, Hadi Sadoghi Yazdi , Hanieh Poostchi:-DT is a call tree that is unremarkably used predictive models in classification. The task of DT is to map observations to focus on values. In DT, every branch representarule. In this paper, it is mainly used to create DTs is planned. It can procedureand execute recursively at every node, till reaching the leaf nodes. It include the shorter height than previous strategies, this can effectively reduce useless variables and also reduce the time required for classification of future data. Unclassified regions also be generated beneath the planned technique, which might be understood disadvantage. The simulation results demonstrate Associate to nursing improvement within the generated call tree compared to previous strategies.

Nguyen Thi Van Uyen, Tae Choong Chung:- We have the tendency to propose a brand new framework for building boosting classifier on distributed databases.At every spherical of the formula, every website processes its own information domestically, and we calculate all required info. That middle website can collect info from all sites and build the new world classifier. This international classifier is mainly employed by every distributed website. The results show that the accuracy of our projected methodology is to sort of capable and accuracy once applying boosting formula to the total dataset. “fast Action Detection via Discriminative Random Forest vote and Top-K Sub volume Search”,Gang Yu,Norberto A. Goussies, Junsong Yuan and Zicheng Liu:- In this project to achieve efficient and robust action detection, we characterize a video of spatio-temporal interest points, and also locate actions for finding spatio-temporal video subvolumes of the highest mutual information score towards each action class. A random forest is constructed. In this project to generate discriminative votes from individual interest points. The detection speed is of magnitude is faster than existing methods.

In this paper, a Patient Treatment Time Prediction (PTTP) model we used historical data. The waiting time of each treatment task is predicted by PTTP, which is the sum of all patients' waiting times in the current queue. Hospital Queue recommendation performs each patient's requested treatment tasks, an efficient and convenient treatment plan with the minimum waiting time for the patient. waiting time of each patient treatment task. It supports large number of data set of every patient and also treatment time predicted for each task of patient. Mobile application can play an important role for giving the information of waiting time of each patient and also information about queue by HQR .PTTP and HQR are based on Random Forest algorithm.Our project is used to reduce waiting time of each patient in hospital.

● It is use to decrease the patients waiting time. ● By using PTTP and HQR algorithm we can helps patients to complete their treatment tasks during a predictable time and helping hospitals schedule every treatment task queue and using this project we can avoid overcrowded and ineffective queues. ● Iterative and KNN algorithm used to display the symptoms and their disease in correct accuracy. ● It is simple and accurate model.

STrain: the training datasets; k: the number of CART trees in the HQR model. Output: PTTP : The HQR system model based on PTTP algorithm Step1: for i ¡- 1 to k do

sampling(STrain); Step 3: create OB subset sOBi ¡-(STrain - straini); Step 4: create an empty CART tree hi; Step 5: for each independent variable yj in straini do Step 6: calculate candidate split points vs ¡- yj; Step 7: for each p in s do Step 8: calculate the best split point (yj,p) ¡- arg min xRL (yi- cL)2+ xRR (yi -cR)2]; Step 9: end for Step 10: append node Node(yj,p) to hi; Step 11: split data for left branch RL(yj,p) ¡- x—yj p; Step 12: split data for right branch RR(yj,p) ¡- xjyj ¿ p; Step 13: for each data R in RL(yj;p); RR(yj;p) do Step 14: calculate 8(vpLjyj) maxi 8(vijy); Step 15: if ((p(L—R)—yj) (p—yj)) then Step 16: append subnode Node(yj,p(L—R)) to Node(yj,p) as multi-branch; Step 17: split data to two forks RL(yj,pL ) and RR(yj,pR); Step 18: else Step 19: collect cleaned data for leaf node Dleaf ¡- (IL yj OL); Step20: calculate mean value of leaf node c 1/K Dleaf ; Step21: end if Step22: end for Step23: remove yj from straini; Step24: end for Step25: calculate accuracy CAi ¡-I (hi(x)=y)/I (hi(x)=y)+I (hi(x)=z) for hi by testing sOBi; Step27: PTTP H(X,j) ¡-1/kiK ¡- [CAi hi]; Step28: return PTTP .

To exhibit the accuracy of our algorithm we conduct some test. We construct nodes. The scales of nodes are placed in the range of (10 to 70). By observing the average execution time of the existing system and proposed system, the execution time is decreases as the number of nodes are increasing. It takes less time than the existing system. The actual operational results of the algorithm are close to the theoretical results.

PTTP and HQR algorithm which is based on Random Forest algorithm .These algorithm is used to predict the time of each patient and also predict each patient waiting time in queue .By using web or android patient can get the information of time taken by each patient. An in this project each patient symptoms and disease is also predicted and also by using mobile application we can get information of ambulance.

G. Biau (2012). ―Analysis of a random forests model,‖ J. Mach. Learn. Res., vol. 13, no. 1, pp. 1063-1095, Apr. 2012. G. Yu, N. A. Goussies, J. Yuan, and Z. Liu (2011). ``Fast action detection via discriminative random forest voting and top-K sub volume search,'' IEEE Trans. Multimedia, vol. 13, no. 3, pp. 507517, Jun. 2011.

Proc. IEEE 7th Int. Conf. e-Business Eng. (ICEBE), Nov. 2010, pp.160-163. K. Farrahi and D. Gatica-Perez (2010). Probabilistic mining of sociogeographic routines from mobile phone data. Selected Topics in Signal Processing, IEEE. Journal of, 4(4): pp. 746-755. K. Farrahi and D. Gatica-Perez (2011). Discovering Routines from Largescale Human Locations using Probabilistic Topic Models. ACM Transactions on Intelligent Systems and Technology (TIST), 2(1). K. Singh, S. C. Guntuku, A. Thakur, and C. Hota (2014). ``Big data analytics framework for peer-to-peer botnet detection using random forests,'' Inf. Sci., vol. 278, pp. 488497, Sep. 2014. Ms.Anuja Bharate, Prof. Devendra Gadekar (2015). ―Survey Paper on Natural Language Processing,‖ IJIEST 2015, vol. No. 1. N. Salehi-Moghaddami, H. S. Yazdi, and H. Poostchi (2011). ``Correlation based splitting criterion in multi branch decision tree,'' Central Eur. J. Comput.Sci., vol. 1, no. 2, pp. 205_220, Jun. 2011. R. Fidalgo-Merino and M. Nunez (2011). ``Self-adaptive induction of regression trees,'' IEEE Trans. Pattern Anal. Mach. Intell., vol. 33, no. 8, pp. 16591672, Aug. 2011. S. Meng, W. Dou, X. Zhang, and J. Chen (2014). ``KASR: A keyword-aware service recommendation method on MapReduce for big data applications,‘‗IEEE Trans. Parallel Distrib. Syst., vol. 25, no. 12, pp. 32213231, Dec. 2014. S. Meng, W. Dou, X. Zhang, and J. Chen (2014). ―KASR: A keywordaware service recommendation method on MapReduce for big data applications,‖ IEEE Trans. Parallel Distrib. Syst., vol. 25, no. 12, pp. 3221-3231, Dec. 2014. S. Tyree, K. Q. Weinberger, K. Agrawal, and J. Paykin (2012). ``Parallel boosted regression trees for Web search ranking,'' in Proc. 20th Int. Conf. World Wide Web (WWW), 2012, pp. 387396. X. Yang, Y. Guo, and Y. Liu (2013). ―Bayesian-inference-based recommendation in online Xiwang Yang, Yang Guo and Yong Liu (2013). Bayesian-Inference-Based Recommendation in Online Social Networks, IEEE 2013. Y.-Y. Chen, A.-J. Cheng, and W. H. Hsu (2013). ―Travel recommendation by mining people attributes and travel group types from communitycontributed photos,‖ IEEE Trans. Multimedia, vol. 15, no. 6, pp. 1283- 1295, Oct. 2013. Yuming Xu,Kenli Li, Ligang He, Longxin Zhang and Keqin Li (2015). A Hybrid Chemical Reaction Optimization Scheme for Task Scheduling on Heterogeneous Computing Systems,IEEE 2015.