OBJECTIVE To better assist drug reference standards management departments in making production scheduling decisions and reduce the impact of stock outs through the construction and application of a drug reference standards sales forecast system. METHODS Based on the sales data of millions of drug reference standards collected from 2008 to 2019, the stock out time of 3682 varieties of national drug reference standards was predicted through data preprocessing and data modeling, then the actual sold-out data in 2020 was used to verify the prediction results, finally the data preprocessing and data modeling were optimized according to the prediction results. RESULTS AND CONCLUSION Through the verification of the actual sales data of reference standards in the four quarters of 2020, the model forecast accuracy rate of the system reaches a high level, and the management personnel of reference materials formulates a more reasonable production scheduling plan with the assist of the model.
胡康, 曹丽梅, 高志峰, 邵俊娟, 路勇, 李健. 销量预测算法在药品标准物质管理中的应用[J]. 中国药学杂志, 2021, 56(16): 1336-1341.
HU Kang, CAO Li-mei, GAO Zhi-feng, SHAO Jun-juan, LU Yong, LI Jian. Application of Sales Forecast Algorithm in the Management of Drug Reference Standards. Chinese Pharmaceutical Journal, 2021, 56(16): 1336-1341.
CAO L M, LIU M L, XIAO X Y, et al. Establishment and introduction of management system of national drug reference standards[J]. Chin Pharm Aff (中国药事), 2017, 31(8):882-886.
[2]
WANG Y J, WANG L, ZHAO Z G. Risk Control of management for standard labels and instructions of national pharmaceutical reference standards[J]. Chin Pharm Aff (中国药事), 2019, 33(4):399-401.
[3]
WANG Q H, SUN L, LIU F, et al. Progress and challenges of reference standards and their new form: digital reference standards (DRS)[J]. Chin Pharm J (中国药学杂志), 2016,51(18):1537-1544.
[4]
MA L Y, LIU M L, MA S C. The development and current situation of national drug reference materials[J]. Chin Pharm J (中国药学杂志), 2012,47(13):1017-1021.
[5]
XIE J J, LIU M L, XIAO X Y, et al. Establishment of comprehensive data platform to guarantee the supply of national drug reference standards[J]. Chin Pharm Aff(中国药事), 2019, 33(1):16-20.
[6]
FITE G, SAM E, TRINQUARD S, et al. PNS66 how accurate are the medicines sales forecasts? a retrospective analysis in France[J]. Value Health, 2020, 23(2):S654-S654.
[7]
MEI X D. Research on influencing factors and prediction of drug sales in retail pharmacies based on machine learning[D]. Beijing: China University of Mining and Tech, 2020.
[8]
ZHOU S Y, XU X, XU X Q. Analysis of carbon dioxide demand forecasting model based on unitary linear regression[J]. Guangdong Chem Ind(广东化工), 2020, 47(23):64-66,75.
[9]
TESHOME H A. Time series analysis of monthly average temperature and rainfall using seasonal ARIMA model (in case of Ambo area, Ethiopia)[J]. Int J Theor Appl Mathematics, 2020, 6(5):76-87.
[10]
HU Y, ZHANG C Y. Prediction of corn price in Hebei province based on ARIMA model[J]. Agric Technol(农业与技术), 2020, 40(23):149-152.
[11]
CAI C Z, WANG H, HE L H, et al. Analysis of world rape yield potential based on ARIMA model[J]. Crop Res(作物研究), 2020, 34(6):511-517.
[12]
CHEN J, WANG X Y, LUO L L, et al. Comparison of machine learning and statistical learning in the imputation of missing values[J]. Stati Decision(统计与决策), 2020, 36(17):28-32.
[13]
MA L M. The choice of outliers in data processing[J]. Dis Surveill(疾病监测), 2010, 25(12):992-993.
[14]
LIE Z H, QUE C Z, XIE Y. Identification and processing method of single outlier of continuous compaction quality measured value[J]. J Harbin Inst Tech(哈尔滨工业大学学报), 2019, 51(3):150-157.
[15]
YU S, GUO Z C, HU L. Treatment of missing data and abnormal data in hospital seasonal time series data[J]. Chin J Health Stati(中国卫生统计), 1990, (6):19-23.
[16]
CHEN Y S. A summary of the processing methods for missing data in time series[J]. China Comput Commun(信息与电脑), 2020, 32(10):19-22.
[17]
QIU H, LI H Y. Research on dynamic data visualization based on Ajax and ecarts[J]. Programming Skills Maintenance(电脑编程技巧与维护), 2020, (10):148-150.
2021, Vol. 56 
