The so-called statistical decision, in a broad sense, is a decision based on statistical principles, principles and methods; In a narrow sense, it refers to the decision-making based on the theory and method provided by probability statistics, and regards the occurrence of future situations as random events. Statistical decision-making provides principles and methods to deal with problems when the future situation is uncertain, and is widely used in enterprise management decision-making.
Basic elements of statistical decision-making
A complete statistical decision-making problem usually contains three basic elements: decision-making goal, natural state and alternative scheme.
Decision goal
Decision-making goal is the goal that decision-makers want to achieve and the starting point and destination of statistical decision-making.
nature
Natural state refers to the objective condition or external environment that is not transferred by the subjective will of the decision-maker, and it is also a factor that affects decision-making.
Alternative
In the decision-making process, there are always two or more options for action.
Application conditions of statistical decision-making
1, quantify the decision target. Statistical decision-making is a hard and quantitative decision, and its decision-making objectives should be quantifiable, such as maximizing profits and minimizing costs.
2. There are more than two (including two) future states, which are also called natural states.
3, more than two (including two) alternative action plans, also known as alternatives, referred to as programs.
4. The return of each action plan in each state should be measurable. The return on income is a measure of the value of the results of the action plan in a given state, and the condition of statistical decision-making is that the results must be computable. In business decision-making, it is generally manifested as some kind of economic profit or loss, such as sales revenue, profit or profit rate.
5. Know the probability of various states, that is, grasp the probability of various states.
The role of statistical decision-making
1. Scientific statistical decision-making plays an intermediary role from decision-making objectives to results.
2. Scientific statistical decision-making provides the best action plan based on facts, which plays a guiding role in avoiding blindness and reducing risks.
3. Statistical decision-making is widely used in many fields such as market, economy and management.
Steps of statistical decision-making
A complete statistical decision-making process must go through the following steps:
1, determine the decision-making target.
Determining the decision-making goal is an important step in decision-making. Without decision-making goals, there is no decision. The so-called decision-making goal refers to the desired result on the basis of prediction under certain environment and conditions. To determine the goal, we must first determine the characteristics and scope of the problem, and then analyze the causes of the problem. At the same time, we should also collect information related to the determination of goals, and then determine a reasonable goal.
Reasonable decision-making objectives can often meet the following requirements:
Accurate meaning, easy to grasp and evaluate;
Quantify the target as much as possible and make clear the time constraint of the target;
Goals should be possible and challenging.
2. Draw up an alternative plan
After the goal is determined, the next task is to analyze the possible ways to achieve the goal, that is, to draw up alternatives. To draw up alternatives, it is necessary to widely collect information related to the decision-making object and environment, and predict various possible ways to achieve the goal and the possible consequences of each way from multiple angles.
Making alternatives is also an innovative process. Policymakers should broaden their thinking, give full play to the role of experience and knowledge, give full play to people's imagination and creativity, and strive to look at decision-making issues from a new angle and a new perspective in order to draw up a novel decision-making plan.
3. Scheme selection
Scheme selection refers to the process of evaluating, comparing and selecting several feasible alternatives to form the optimal action scheme. In the evaluation and analysis, it is necessary to determine the evaluation elements, evaluation criteria and evaluation methods of the scheme according to the predetermined decision-making objectives and established value standards, and sometimes some sensitivity analysis is needed. In addition, when choosing a scheme, besides the prototype of the alternative scheme, it can also be a modified scheme of one scheme or a new scheme obtained by combining several alternatives. When conditions permit, the evaluation process should try to carry out typical tests on relevant schemes or use computers to carry out simulation tests.
4. Programme implementation
After the plan is determined, it is necessary to organize manpower, material resources and financial resources to implement the decision-making plan. In the process of decision-making, the decision-making body must strengthen supervision, feed back the information of the implementation process to the decision-makers in time, and take measures to correct the deviations in time. If the implementation of the decision is unexpected, or the environmental state changes greatly, the implementation of the decision should be suspended, the decision-making objectives and decision-making scheme should be re-examined, and the objective situation should be adapted by modifying the objectives or replacing the decision-making scheme. The implementation should be flexible. The statistical decision-making process is as follows: