"... In PAGE 4: ...isk for outbreaks, 1993-1995 .................................................................... 47 Tables Table1 : Yellow fever vaccinations with French neurotropic vaccine and cases of yellow fever in Africa, 1934-1953 .... In PAGE 4: ...frican countries at risk for yellow fever outbreaks................................ 49 Table1 0: Prioritising 34 African countries at risk for yellow fever for support; highest priority.... In PAGE 4: ...ighest priority............................................................................................... 50 Table1 1: Prioritising 34 African countries at risk for yellow fever for support; high priority.... In PAGE 4: ...igh priority.................................................................................................... 51 Table1 2:Prioritising 34 African countries at risk for yellow fever for support; medium priority .... In PAGE 4: ...edium priority ............................................................................................. 52 Table1 3: Prioritising 34 African countries at risk for yellow fever for support;... In PAGE 20: ... Milestones in the development and use of French neurotropic vaccine are summarized in the box. Between 1939 and 1952 over 38 million doses were administered (mostly by scarification along with smallpox vaccine) in Francophone countries of West Africa, and incidence declined dramatically ( Table1 ). However, a high incidence of encephalitic reactions in children led to its use in children under 10 years being stopped in 1961, and manufacture of the vaccine was discontinued in 1980.... In PAGE 21: ... 1941: A departmental order made yellow fever vaccination by scarification compulsory for the whole civilian and military population of French West Africa. Yellow fever virtually disappeared from colonial French West and Equatorial Africa by virtue of a programme of compulsory immunization initiated in 1942 ( Table1 ). The same period was marked by major epidemics in the British colonies of Gold Coast and Nigeria, which had not implemented a policy of preventive immunization.... In PAGE 23: ...37, 38 Before mass immunization campaigns were started in Africa, typical urban outbreaks occurred in Lagos, Nigeria, in 1925-1926, in Accra, Ghana in 1926-1927 and again in 1937, and in Banjul , the Gambia, in 1934-1935.10 In 1940, mass immunization was initiated in French-speaking countries in West and Equatorial Africa where 25 million people were immunized about every four years ( Table1 ). As a consequence, yellow fever disappeared gradually in these countries, while epidemic and endemic... In PAGE 58: ...Yellow fever Table1 0: Prioritising 34 African countries at risk for yellow fever for support; highest priority Highest priority Country Arguments Comments Recommendations Nigeria - huge epidemics during - has already included - yellow fever vaccination recent years: highest YF in the EPI, but together with measles number of reported coverage only 1% at the age gt; 6 months cases in Africa - the whole country - centre of other should be included epidemics in topotype II - improve surveillance area Cameroon - neighbouring Nigeria - 214 000 doses of YF - as Nigeria - two epidemics in 1990s vaccine given in 1990 Kenya - topotype III in Kerio - mass vaccinations in - yellow fever vaccination Valley 1992-1993 covered needs to be included - has established one million people into the EPI in Kerio sentinel surveillance for Valley, also all yellow fever immigrants should be - endemic area close to vaccinated. big highly populated - existing surveillance urban centres system needs to be improved Liberia - CIGN*** - mass vaccinations - yellow fever vaccination - most recent epidemic covered one million together with measles reported in Africa (1997) people in 1995 at the age gt; 6 months - improve surveillance Mali - CIGN*** - mass vaccination - EPI activities need - centres the topotype I campaign in 1969, support area and three million were - YF vaccination should - reported one epidemic vaccinated in 1987 be included in EPI in in 1987; 70% lt; 15 years high risk areas old - improve surveillance ** No reports to WHO for the last three to five years.... In PAGE 59: ...WHO/EPI/GEN/98.11 Table1 1: Prioritising 34 African countries at risk for yellow fever for support; high priority High priority countries Country Arguments Comments Recommendations Angola - reported an epidemic in - has already included - only part of the country 1988 YF in the EPI needs YF vaccinations, but the overall EPI programme needs to be improved - support surveillance Burkina - CIGN - has already included - support EPI activities Faso - reported an epidemic in YF in the EPI - support surveillance 1985 Gabon - reported an epidemic in - has already included - YF immunization 1995 YF in the EPI coverage poor, surveillance system needed Mauritania - CIGN - has already included - only part of the country - reported one epidemic YF in the EPI needs YF vaccinations in 1987 - but connected with Malian epidemic Senegal - reported an epidemic - has already included - support EPI and in 1995 YF in the EPI surveillance to sustain - high risk: historically - measles immunization good control many epidemics coverage 80%, but YF - monitor YF vaccination immunization Coverage coverage and missed lt;50% opportunities Togo - reported an epidemic in - Difficult to assess - Introduce yellow fever 1987 degree of risk vaccination together with measles at the... In PAGE 60: ...Yellow fever Table1 2: Prioritising 34 African countries at risk for yellow fever for support; medium priority Medium priority countries Country Arguments Comments Recommendations Benin - reported an epidemic in - measles immunization - training needed to add 1996 coverage good YF vaccinations to EPI - historically many epidemics CAR - CIGN*** - has already included - support EPI activities - no reported epidemics YF in the EPI - support surveillance for tens of years Chad - CIGN*** - has already included - support EPI activities - no reported epidemics YF in the EPI - support surveillance for tens of years Congo - last epidemic in 1961 - support EPI activities - support surveillance Eq. Guinea - CIGN*** - support surveillance - last cases in 1970 Ethiopia - CIGN*** - support EPI activities - last epidemic in 1966 - support surveillance Ghana - reported the second - has already included - support surveillance highest number of YF in the EPI cases during the past - YF coverage gt; 50% 15 years Guinea - last cases in 1987 - support EPI activities - support surveillance Ivory Coast - last epidemic in 1982 - has already included - support surveillance YF in the EPI and EPI activities Niger - CIGN*** - has already included - support surveillance - last epidemic in 1939 YF in the EPI and EPI activities Sierra - CIGN*** - include yellow fever in Leone - reported an epidemic in the EPI 1995 Sudan - last epidemic in 1942 - Part of country at risk - include yellow fever in the EPI, but only partially Uganda - last reported cases - support surveillance from 1971 Former - CIGN*** - support surveillance Zaire - last epidemic in 1972 and EPI activities ** No reports to WHO for the last three to five years.... In PAGE 61: ...WHO/EPI/GEN/98.11 Table1 3: Prioritising 34 African countries at risk for yellow fever for support; lowest priority Lowest priority countries Country Arguments Comments Recommendations Burundi - CIGN*** - surveillance and - no reported epidemics outbreak response Cape - no reported epidemics - surveillance and Verde Is. outbreak response Eritrea - CIGN*** - surveillance and - no reported epidemics outbreak response Gambia - good immunization - has already included - support surveillance performance: no YF in the EPI reported epidemics since 1979 Guinea - no reported cases since - surveillance and Bissau 1951 outbreak response Rwanda - no reported epidemics - surveillance and outbreak response Sao Tome - no reported epidemics - surveillance and amp; Principe outbreak response Somalia - CIGN*** - surveillance and - no reported epidemics outbreak response Tanzania - no reported epidemics - surveillance and (even historical)... ..."

"... In PAGE 7: ...o the relation given in Focks et al. [24]. For each municipality, we estimate a mean temperature using the monthly temperature measures during the exponential phase of the epidemic. The esti- mated mean temperatures ( Table1 ) are above the minimal hatching temperatures for mosquito eggs (varying from 20 to 13 C176C) obtained from experimental studies [11]. The mosquito mortality rate is assumed to be independent of ambient temperature as in other studies [45] and sampled from a Gamma distribution with a mean of 10.... ..."

"... In PAGE 4: ...able 4: Epidemics reported 1984-1996 ................................................................... 26 Table5 : Ecological factors affecting yellow fever transmission .... In PAGE 32: ...II.5. Recent epidemiology in Africa The period 1986-1991 was an extraordinarily active period for yellow fever in Africa. The world wide total of 20 424 reported cases and 5447 deaths represented the greatest yellow fever activity reported to WHO since reporting began in 1948 ( Table5 / Map 3).53, 54, 55 The largest number of cases was reported from Nigeria, where a resurgence of yellow fever has been noted since 1984.... ..."

"... In PAGE 7: ...rates (Ka and Ks) for the ECD, ICD, and kinase domains of each pair (Figures 6B, 6C, and 6D, respectively). Interestingly, the frequency distribution and the mean of the Ka/Ks ratios for ECDs are significantly different from those of ICDs and kinase domains ( Table2 ), indicating that the ECDs evolved faster than the ICDs. This difference may either be because of relaxed purifying selection or positive selection on the ECDs.... ..."

"... In PAGE 21: ... of Control 68 (1997): 259-76. Table1 . Partial List of BIO-Plex Devices INVESTIGATIVE TEAM UH PI: G.... In PAGE 50: ...ig. 1. The Q-Q plot and the CDF plot for soak time and cycle time for uncensored observations. Table1 . Predictions made by Logistic and Weibull Models Logistic Model Predictions Weibull Model Predictions Survive Fail Survive Fail Actually Survived 84 36 35 85 Actually Failed 31 75 26 80 times.... In PAGE 54: ... Air samples were screened for a specific set of volatiles to ensure that they did not exceed Spacecraft Maximum Allowable Concentrations.6 Table1 shows the average and stan- dard deviation of some of the accumulated volatiles for three soymilk production runs. The grinding por- tion of the cycle contributed only minor amounts of ethanol (0.... In PAGE 54: ...37 percent of total ethanol detected). These compounds ( Table1 ) likely resulted from lipid oxida- tion and/or Maillard reactions during heating.7 To determine the accumulation of these compounds during a long-duration mission (180 days) for a crew of four, a hypothetical diet was used, in which soymilk was consumed as a beverage, as well as an ingredient in other items.... In PAGE 56: ...that SoyaCowTm would reside in the tunnel,8 which has a volume of 202m3. Table1 shows 180-day SMAC values for the different volatiles as well as for total quantities expected to be generated during soymilk production in the tunnel (assuming no leaks to the other chambers in the ALSSITB). It is apparent from Table 1 that the SMAC values for all the com- pounds would be exceeded during the 180-day period if no appropriate trace contaminant control system were used.... In PAGE 56: ... Table 1 shows 180-day SMAC values for the different volatiles as well as for total quantities expected to be generated during soymilk production in the tunnel (assuming no leaks to the other chambers in the ALSSITB). It is apparent from Table1 that the SMAC values for all the com- pounds would be exceeded during the 180-day period if no appropriate trace contaminant control system were used. For example, acetaldehyde would exceed the SMAC value by 1,841 times and would result n an accumulation of 1,488 grams in the tunnel.... In PAGE 56: ...9 *A/B ratio signifies the multiple of each volatile compared to its SMAC value. Table1 . Quantification of volatiles with Spaceraft Maximum Allowable Concentrations (SMAC) evolved from the soymilk preparation process.... In PAGE 69: ...0 in. wide, and 1/32 in. to 1 in. thick. The informa- tion in the Connector Material Suggested column from Table1 indicates that a side-w may be connect- ed with other side-w components using a metal con- nector or a chemical compound connector. This infor- mation is used by the ADL generated to restricts the connectors available to the designer to those defined in Connector Material in Table 2.... In PAGE 69: ... This infor- mation is used by the ADL generated to restricts the connectors available to the designer to those defined in Connector Material in Table 2. Each of the sides may be an instance of the compo- nents available in Table1 (side-w, side-m, side-c, side-p). For example, a side-w component comple- ments the domain rules with information such as Component Material, Available Dimensions, Connector Material Suggested, and C o m p o n e n t Table 1.... In PAGE 70: ...Properties contained in Table1 . The domain rules also indicate the connection rules that dictate the manner in which the borders (b1, b2, b3, b4) of the components will be connected to the borders of other components using a connector (nail, glue, screw, sta- ple).... In PAGE 78: ... Development of the AUTO-ID capability represents a significant technol- ogy advancement. Table1 provides X-38 modal fre- quency and modal damping results obtained using AUTO-ID. The data set was produced during a cap- tive carry test as shown in Fig.... In PAGE 78: ...77 1.9 Table1 . X-38 Captive Carry AUTO-ID Results Fig.... ..."

"... In PAGE 17: ... If the values of each attribute domain are disjoint then the columns of the example relations are disjoint; we therefore take the domain size for the relation to be the maximal domain size in any one column. Table20 depicts the nal state of an evolutionary run applied to this set of dependencies. It is an Armstrong relation but has only six tuples and a domain size of 4.... ..."

"... In PAGE 7: ... If the values of each attribute domain are disjoint then the columns of the example relations are disjoint; we therefore take the do- main size for the relation to be the maximal domain size in any one column. Table1 depicts the nal state of an evolutionary run applied to this set of dependencies. It is an Armstrong relation but has only six tuples and a domain size of 4.... ..."

"... In PAGE 3: ...1 For the schema R containing attributes TEACHER, HOUR, and ROOM we impose the FD TEACHER HOUR ! ROOM, implying that no teacher can be give a class in more than one room at the same time. We denote TEACHER, HOUR, and ROOM by T; H and R, respectively, and show a deterministic generated Armstrong relation (AR) in Table 1, an evolved Armstrong relation in Table2 (with an equivalent domain and one less tuple), an evolved relation in Table 3 with three tuples depicting the best attainable quality (0.5) for the FD in a relation of this size, and in Table 4 we provide a counterexample relation for the FD ROOM ! HOUR which implies that a room can only ever be used at one time whilst satisfying our original speci ed FD.... In PAGE 18: ...with 9 tuples and a domain size 5. A B C D E 0 0 0 0 0 1 0 1 1 1 2 1 1 2 2 2 2 1 3 3 2 3 1 4 3 2 3 2 4 3 2 4 2 4 3 Table 19: Standard AR A B C D E 1 0 3 0 0 1 1 3 1 1 1 2 3 2 0 1 3 2 2 0 1 3 3 2 0 2 1 2 3 2 Table2 0: An evolved AR A B C D E 2 3 0 4 3 4 1 0 1 1 4 1 2 1 1 4 3 2 0 4 Table 21: Evolved relation A B C D E 0 1 2 4 0 0 2 0 0 4 0 4 0 4 0 2 2 3 2 1 3 0 1 3 2 3 1 4 1 2 3 2 1 1 2 3 2 4 1 2 3 3 2 1 2 Table 22: A larger AR The evolutionary procedure is now highlighted with some real world examples, with the in- tention of providing the reader with an appreciation of the utility of varied example relations. It can be said that a greater understanding of the semantics of an FD set is reached by repeated examinations of di erent instances of the example relations; this is one motivating factor behind our probabilistic approach.... In PAGE 18: ...with 9 tuples and a domain size 5. A B C D E 0 0 0 0 0 1 0 1 1 1 2 1 1 2 2 2 2 1 3 3 2 3 1 4 3 2 3 2 4 3 2 4 2 4 3 Table 19: Standard AR A B C D E 1 0 3 0 0 1 1 3 1 1 1 2 3 2 0 1 3 2 2 0 1 3 3 2 0 2 1 2 3 2 Table 20: An evolved AR A B C D E 2 3 0 4 3 4 1 0 1 1 4 1 2 1 1 4 3 2 0 4 Table2 1: Evolved relation A B C D E 0 1 2 4 0 0 2 0 0 4 0 4 0 4 0 2 2 3 2 1 3 0 1 3 2 3 1 4 1 2 3 2 1 1 2 3 2 4 1 2 3 3 2 1 2 Table 22: A larger AR The evolutionary procedure is now highlighted with some real world examples, with the in- tention of providing the reader with an appreciation of the utility of varied example relations. It can be said that a greater understanding of the semantics of an FD set is reached by repeated examinations of di erent instances of the example relations; this is one motivating factor behind our probabilistic approach.... In PAGE 18: ...with 9 tuples and a domain size 5. A B C D E 0 0 0 0 0 1 0 1 1 1 2 1 1 2 2 2 2 1 3 3 2 3 1 4 3 2 3 2 4 3 2 4 2 4 3 Table 19: Standard AR A B C D E 1 0 3 0 0 1 1 3 1 1 1 2 3 2 0 1 3 2 2 0 1 3 3 2 0 2 1 2 3 2 Table 20: An evolved AR A B C D E 2 3 0 4 3 4 1 0 1 1 4 1 2 1 1 4 3 2 0 4 Table 21: Evolved relation A B C D E 0 1 2 4 0 0 2 0 0 4 0 4 0 4 0 2 2 3 2 1 3 0 1 3 2 3 1 4 1 2 3 2 1 1 2 3 2 4 1 2 3 3 2 1 2 Table2 2: A larger AR The evolutionary procedure is now highlighted with some real world examples, with the in- tention of providing the reader with an appreciation of the utility of varied example relations. It can be said that a greater understanding of the semantics of an FD set is reached by repeated examinations of di erent instances of the example relations; this is one motivating factor behind our probabilistic approach.... In PAGE 19: ...Phone Flat no. Postcode City Dave 1246 19 NW1 London Dave 1246 19 YO2 York Dan 3748 7 YO2 York Dan 3748 7 YO1 York Charles 3748 11 YO1 York Table2 3: Mannila apos;s Armstrong relation for set 29 Name Phone Flat no. Postcode City Dave 1246 19 NW1 London Dave 1246 19 YO2 York Dan 3748 7 NW1 London Dan 3748 7 W14 London Charles 1246 11 YO2 York Table 24: An evolved AR with the same domain size j GEN(F) j in the size of the FD set are discussed in [BDFS84].... In PAGE 19: ... Postcode City Dave 1246 19 NW1 London Dave 1246 19 YO2 York Dan 3748 7 YO2 York Dan 3748 7 YO1 York Charles 3748 11 YO1 York Table 23: Mannila apos;s Armstrong relation for set 29 Name Phone Flat no. Postcode City Dave 1246 19 NW1 London Dave 1246 19 YO2 York Dan 3748 7 NW1 London Dan 3748 7 W14 London Charles 1246 11 YO2 York Table2 4: An evolved AR with the same domain size j GEN(F) j in the size of the FD set are discussed in [BDFS84]. These sets possess the following pattern: fA1A2 ! B; A3A4 ! B; : : :; A2m?1A2m ! Bg which has j GEN(F) j= 2m and for our simulations resulted in examples with a low proximity to Armstrong.... In PAGE 20: ...Phone Flat no. Postcode City Dave 1246 19 YO1 York Dave 1246 19 NW1 London Dan 1246 7 W14 London Dan 1246 7 NW1 London Charles 3748 11 YO2 York Table2 5: An AR with a larger domain size Name Phone Flat no. Postcode City Dave 1246 19 NW1 London Dave 1246 19 W14 London Dave 1246 19 YO3 York Dan 1246 7 BS8 Bristol Dan 1246 7 BA1 Bath Dan 1246 7 BA2 Bath Charles 1246 11 YO2 York Matt 8881 84 BA8 Bath Fred 2383 24 YO3 York Table 26: An evolved AR with 9 tuples For pathological sets an Armstrong relation could be evolved with a higher probability if the number of tuples were initially much larger than the number of elements in GEN(F) and the domain size were small, as the likelihood of providing counterexamples within such an example increases.... In PAGE 20: ... Postcode City Dave 1246 19 YO1 York Dave 1246 19 NW1 London Dan 1246 7 W14 London Dan 1246 7 NW1 London Charles 3748 11 YO2 York Table 25: An AR with a larger domain size Name Phone Flat no. Postcode City Dave 1246 19 NW1 London Dave 1246 19 W14 London Dave 1246 19 YO3 York Dan 1246 7 BS8 Bristol Dan 1246 7 BA1 Bath Dan 1246 7 BA2 Bath Charles 1246 11 YO2 York Matt 8881 84 BA8 Bath Fred 2383 24 YO3 York Table2 6: An evolved AR with 9 tuples For pathological sets an Armstrong relation could be evolved with a higher probability if the number of tuples were initially much larger than the number of elements in GEN(F) and the domain size were small, as the likelihood of providing counterexamples within such an example increases. A better technique, however, would be to incorporate a penalising procedure which violates dependencies that hold in a relation but which are outside of the speci ed set.... ..."

"... In PAGE 20: ... Therefore, these country-by-country regressions used a wealth index that was recalculated country by country. Table9 reports the selected p-values of the joint tests of significance of the wealth and cluster-level wealth and fever variables in a model that also includes all the other control variables (see Table 8) and that allow all the coefficients to differ across countries. Table 9.... ..."