With the layer multinet model proposed by our research group, we investigated task-based algorithms for constructive neural networks, their perspective, strength and weakness 還結(jié)合所提出的層次式多網(wǎng)絡(luò)模型,對(duì)基于任務(wù)的神經(jīng)網(wǎng)絡(luò)的自適應(yīng)結(jié)構(gòu)方案、發(fā)展前景與存在的問(wèn)題作了詳細(xì)的比較研究。