1: \documentclass[12pt,english]{article}
2: \usepackage[cp866]{inputenc}
3: \usepackage[english]{babel}
4: \usepackage{amsmath}
5: \usepackage{graphicx}
6: \topmargin -1,5cm \textheight 24cm \textwidth 17cm \oddsidemargin
7: 0cm
8: \begin{document}
9: \begin{center}
10: {\Large\bf Factor Temporal Prognosis \\
11: of Tick-Borne Encephalitis Foci Functioning on the South of
12: Russian Far East}
13: \medskip
14:
15: {\bf E.I. Bolotin $\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;$G.Sh. Tsitsiashvili\\
16: Pacific Ocean Institute of Geography $\;\;\;\;\;\;\;\;\;$ Institute of Applied Mathematics\\
17: Far Eastern
18: Branch of RAS $\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;$ Russia Far Eastern Branch of RAS\\
19: Vladivostok, Russia $\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;$ Vladivostok, Russia\\
20: .$\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;
21: \;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;$e-mail:guram@iam.dvo.ru\\
22: I.V. Golycheva \\
23: Institute of Applied Mathematics\\
24: Far Eastern Branch of RAS Vladivostok Russia\\
25: e-mail: iragol@yandex.ru }\end{center}
26: \medskip
27: \begin{center}
28: {\large\bf Abstract}
29: \end{center}
30:
31: A method of temporal factor prognosis of TE (tick-borne
32: encephalitis) infection has been developed. The high precision of
33: the prognosis results for a number of geographical regions of
34: Primorsky Krai has been achieved. The method can be applied not
35: only to epidemiological research but also to others.\\\\
36:
37: {\large\bf 1. Introduction}
38:
39: While studying various complex systems the examination of their
40: functioning both in the past and in the future is of great
41: importance. The most informative index of natural pest foci is the
42: changing of human infection rate. Epidemic temporal rows of
43: infection integral expression of influence of a number of factors.
44: They comprise prognosis data which can be used to study this or
45: that phenomenon in future [1].
46:
47: Prognosis approach in scientific research, including the
48: development of methodological base of natural - foci infection
49: rate prognosis, is the main research and practical goal. It is of
50: great importance nowadays because the epidemic situation has
51: become worse not only in Russia, but also in many regions and
52: countries of the world.
53:
54: There exist two ways of temporal prognosis. The first one is
55: extrapolation prognosis. It is based on the analysis of temporal
56: rows of infection for the past several years and on their direct
57: extrapolation.
58:
59: The second approach to temporal prognosis is a factor one and it
60: has several stages. Possibilities and advantages of temporal
61: factor prognosis are the following.
62:
63: $\bullet$ First, a researcher, using extrapolation prognosis,
64: deals with average characteristics which reflect general trend
65: of phenomenon development. This trend is an artificial line where
66: the most important for prognosis of critical level conditions of
67: the analysed system are graded.
68:
69: $\bullet$ Second, the results of the extrapolation prognosis
70: depend on the length of temporal raw and are specified by the type
71: of long standing development of the researched phenomenon.
72:
73: $\bullet$ Third, the results of extrapolation prognosis have
74: probable character and are limited by confidence intervals. The
75: range of such limits, however, can be so broad that it makes the
76: obtained results so "vague" that considerably limits possibility
77: of their objective interpretation.
78:
79: $\bullet$ Forth, extrapolation prognosis is realized by the "black
80: box" principle, it is deprived of the possibility to analyse the
81: caused factor and mechanism which form the temporal raw.
82:
83: These limitations define our interest to more sophisticated and
84: yet more objective factor method in the prognosis of the epidemic
85: activity of tick-borne encephalitis, the area of which extends
86: from Europe to Japan [2], [3].
87:
88: The levels of connection turned out to be unstable according to
89: what we have found out from our special preliminary research in
90: study of connection stability between the long standing row of
91: infection rate and some influential factors. Such unstable
92: correlation between infection rate and influential factors
93: strictly limits the possibilities of linear statistic models use.
94: Thus, the prognosis goal setting was changed. The nature of which
95: was to forecast indefinite index but infection rates which could
96: be higher or equal to some extreme line given by an expert. Such
97: goal setting is particularly important methodologically because
98: the way of its realisation has a unique character and can be
99: easily reproduced in other scientific spheres.
100:
101: The factor prognosis research which we use has been divided into
102: two stages. The first stage included the realisation of the
103: epidemiological index and the questions concerning the potential
104: prognosis results accuracy by the different statistic selection
105: were also discussed. The aim of the second research stage was
106: firsthand factor prognosis of critical infection rate of tick-bome
107: encephalitis in a real time. Thus, while at the first stage of the
108: investigation purely academic problems were solved and
109: methodological approaches of putting it into practice were
110: developed, the second stage can be of practical importance not
111: only for tick-bome encephalitis but for wide range of other
112: phenomena as well.\\\\
113:
114: {\large\bf 2. Main results of research}
115:
116:
117: The experimental material for factor temporal prognosis presents
118: statistic data on tick-bome infection dynamics during the last 30
119: years in a number of epidemic areas of Primorski Krai, previously
120: determined by us [4].
121:
122: On prognosis on the first investigational stage the use was made
123: of influential factors that are long standing rows of average
124: annual meaning of 8 meteorological factors of 3 meteorological
125: stations, that characterise southern, middle and northern regions
126: of Primorski Krai. The used meteorological factors include:
127: average January temperature, absolute temperature minimum, average
128: annual temperature, the length of non-frosty period, average May
129: temperature, average snow blanket and snow-temperature
130: coefficient.
131:
132: On the second stage of temporal factor prognosis the materials on
133: long standing dynamics of population Ixodes persulcatus - the main
134: TE virus carrier were drawn in addition.
135:
136: Methods of factor prognosis of critical levels of TE infection
137: were based on the following algorithm worked out by us. Firstly,
138: empirical information of TE and influential factors was introduced
139: as the matrix of initial data in which rows indicate years and
140: columns indicate index of infection and influential factors.
141: Further, the years rows with critical level of infection and
142: corresponding indexes of influential factors were emphasized. The
143: emphasized years with critical levels of infection form intervals
144: of particular influential factor meanings. However, these
145: intervals may also include years in which the infection is lower
146: than critical level that we emphasized earlier. Let's mark such
147: years as "false critical". Having marked the amount of critical
148: and "false critical" years as "$x$" and "$y$" accordingly. We have
149: a ratio $p=x/(x+y)$ which can be interpreted as possibility to
150: identify correctly the critical years on emphasized intervals of
151: influential factors.
152:
153: As a whole the developed method of factor temporal prognosis was
154: based on the image recognition principle. But the quality
155: prognosis, realized by original algorithm, was determined
156: according to selection by number of non-critical years falsie
157: taken for critical by the determined final rule of recognition.
158: And at the first stage of investigation the final rule of the year
159: recognition as a critical was based on the condition of 100$\%$
160: this year being in the defined meaning intervals of influential
161: factors of critical years.
162:
163: However, at the second stage of investigation i.e. on prognosis of
164: critical events in real time, it turned out that such condition is
165: just of secondary meaning. This conclusion was made during
166: numerous calculating experiments that we conducted.
167:
168: As a result, new prognosis rule was established in which a number
169: of identifying factors fluctuated between 50 and 75 percents of
170: all the used ones. Critical infection rate was also chosen
171: depending on statistic sample with the amount of critical years
172: being more that one. It should be noted that during the repeated
173: besting of prognosis rule, simplicity and speed of included
174: calculations were of the primary importance.
175:
176: On the whole, some most significant moments of conducted factor
177: prognosis can be signed out. Thus, accuracy of factor prognosis
178: totalled from 50 to 100 percents.
179:
180: The maximum prognosis rate was provided southerner territories of
181: investigated region, however, some other geographic regions
182: characterized by close prognosis figures were mentioned.
183:
184: The prognosis rate is considerably influenced by the amount and
185: concrete set of affecting factors while the increase of used
186: factors number does not always result in higher prognosis
187: accuracy. This remarkable fact requires further thorough
188: investigation and specification.
189:
190: The received data concerning the connection between prognosis
191: accuracy and used critical infection rate are also interesting. It
192: was found out that in some cases this connection can be direct and
193: indirect in the others. This quite substantial fact also requires
194: serious analysis and consideration.
195:
196: It was revealed that the usage of factual materials on long-term
197: Ixodes perculcatus's population dynamics had not improved
198: prognosis accuracy. It remained the same or even lowered. This
199: tendency again proves our opinion about the insignificant
200: influence of carriers population on human's suffering from
201: tick-borne Encephalitis. If we compare the results of both the
202: first and second stages of temporary factor prognosis, their
203: considerable similarity can be marked.\\\\
204:
205: {\large\bf 3. Conclusion}
206:
207: On the whole, the developed method of critical infection rates
208: temporary factor prognosis solves an extremely vital problem of
209: non-linear .character, from the other side it has several
210: important features.
211:
212: $\bullet$ First, this method is quite easy for realization, but at
213: the same time it is all-purpose; namely it is able to work with
214: any information presented in the form of dynamic rows.
215:
216: $\bullet$ Second, all the calculations performed with the aid of
217: prepared program can easily be conducted manually.
218:
219: $\bullet$ Third, the proposed method can be regarded as one of the
220: highest level steps for real prognosis , based on dynamic
221: cause-and-effect models.
222:
223: $\bullet$ Forth, the method of critical disease rate analysis has
224: good perspectives as it deals with a number of important
225: problems. For example, how will the results of the prognosis
226: change if the number and nature of the factors are altered? What
227: is the effect of modifying the length of rows and critical disease
228: ranges? How can the quality of the program of the prognosis be
229: modified in case of correlating the disease and the influential
230: factors with different logs? How is the influential factors
231: hierarchy modified with regard to their prognosis importance in
232: the course of temporal rows analysis? There also exits quite a
233: number of vitally important problems. This work is the first and
234: quite a modest attempt to solve some of these problems.
235:
236: Thus, the developed temporal factor prognosis method being of both
237: theoretical and practical interest, poses a number of important
238: questions as well. Having solved the problems by means of
239: wide-range testing of the proposed algorithm, it is possible to
240: receive, for the future, a universal method of temporal factor
241: prognosis.\\\\
242:
243: {\large\bf References}\\
244: 1. Iagodinsky V.N. "Dynamics of Epidemiological Process", Moscow,
245: Medicine, 1977, 240 p.\\
246: 2. Zlobin V.I., Gorin O.Z. "Tick-Borne Encephalitis", Novosibirsk,
247: Medicine, 1996, 177 p.\\
248: 3. Korenberg A.I. "Biohorological Structure of Species (on Example
249: of Taiga Tick)", Moscow, Science, 1979, 172 p.\\
250: 4. Bolotin E.I. "On a Functional Organization of Natural Tick-Bome
251: Encephalitis Foci and Prognosis of Their Epidemic Manifestation:
252: Analysis of Monodimensional Temporal Rows of Infection".
253: Parazitology, 2001; 35, 5, 386-393.
254:
255:
256:
257:
258:
259: \end{document}
260: