中国与吉尔吉斯斯坦边界演变

中国与吉尔吉斯斯坦的边界问题是沙俄侵略中国西北边疆的产物，历经中俄、中苏、中国与独立的吉尔吉斯斯坦的漫长谈判，最终得以解决。由于清朝政府在边界问题上的失误，使得问题复杂化。虽然中吉解决了边界问题，但此问题时不时被拿来用作其国内政治斗争的工具和反华、排华的招牌。     

阿合奇县哈拉奇乡柯尔克孜族信仰状况调查

文章通过在阿合奇县哈拉奇乡实地问卷调查研究,指出柯尔克孜族宗教信仰有以下特点：(1)柯尔克孜族宗教氛围不浓厚;(2)牧区宗教意识相对淡薄,清真寺数量较少,场所简陋,其宗教教职人员相对较少;(3)游牧的生活方式不允许他们完全履行伊斯兰教教义规定的一切义务,牧民受伊斯兰教的影响也因此受到限制;(4)宗教从来没有左右柯尔克孜族,伊斯兰教在柯尔克孜族中处于一种相对自由的状态。     

新疆塔城柯尔克孜族的藏传佛教信仰特征

新疆塔城地区大多柯尔克孜人信仰藏传佛教,这个特殊的族群由于被动接触,缺乏对藏传佛教的深刻体会和情感联系.其信仰特征如下：1.对藏传佛教知识的掌握程度不高,但普遍对藏传佛教有深厚的情感;2.藏传佛教给塔城信仰此教的柯尔克孜族提供了安全感和归属感;3.随着时代的变迁,他们对佛的观念日趋淡化,宗教已是世俗化趋势;4.宗教融汇于其生活和风俗习惯之中,并将长期存在.     

Generalized functions in the qualitative study of heterogeneous populations

ABSTRACT

Solutions from non-smooth functional spaces, including generalized functions and measures, often appear in optimal control theory but are avoided in applications. They are however useful in finding the optimal distribution of investments into new and old capital equipment under improving technology. The corresponding economic problem involves optimal control in a linear Lotka-McKendrik model of age-structured population. Optimal solutions do not exist in normal functional classes and, so, generalized functions are used to construct the solutions. The optimal age-distributions of capital and investment include the Dirac function and are interpreted as instantaneous investment in equipment of certain age. A numerical simulation completes the presentation of the dynamics.     

Untransformed first observation problem in regression model with moving average process

This paper shows the impact of underestimation of variance of an estima-tor when first observation is left untransformed to simplify the computational procedure. In fact, the bias of the variance is not diminishing even for large sample size for the model considered. By partitioning the covariance matrix into two parts, this paper explains why least square estimator with untrans-formed first observation shows such a consequence. To demonstrate this, an exact GLS estimator is developed by modifying an approximate estimator. Nonetheless, the computational simplicity remains same.